Skip to main content

Full text of "The New Astronomy"

See other formats


& Lesley 

Paul & Lesley Murdin 


black holes, white dwarfs, pulsars, quasars and supernovae — 
how radio astronomy is changing our concepts of the universe 

Paul and Lesley Murdin 

&/rtc£ Aofej white dwarfs, pulsars, 
quasars and supernovae- 
how the new astronomy % 
our concepts of the universe 



^ ••!.>-; 

Reference International 
London • New York 

The Crab Nebula — arguably the most important object in the sky to the study 
of astrophysics. Red filaments of hydrogen gas still glow more than goo years 
after Chinese astronomers saw a brilliant "guest star" appear in this spot. 

© 1978 Reference International Publishers Ltd. 

First published in UK 1973 

by Reference International Publishers Limited 

21 Soho Square, London Wi 

All rights reserved. No part of this publication 
may be reproduced or transmitted, in any form 
or by any means, without written permission 
from the publisher. 

isbn o 905154 06 1 

To our parents 

That thd' a man were admitted into heaven 
to view the wonderful fabrick of the world, 
and the beauty of the stars, yet what would 
otherwise be rapture and extasie, would be 
but a melancholy amazement if he had not a 
friend to communicate it to. 

Archytas (attr.) 


I Supernovae in space and time i 

II Guest Stars: the historical supernovae 6 

III Renaissance supernovae: 

shattering the crystal spheres 25 

IV Supernovae in other galaxies: supernovae sought 50 
V The Crab and its mysteries: a supernova remnant 66 

VI Discovering pulsars: 

heartbeats of supernova remnants 97 
VII The search for supernova remnants: 

looking for other Crabs 118 

VIII Types of supernovae: gathering the evidence 131 
IX The making of a neutron star: 

what makes pulsars tick 144 
X Neutrino astronomy: 

the ultimate cause of supernovae 155 
XI Creation of the elements: 

Man, the supernova remnant 161 

XII Cosmic rays: supernovae and evolution 177 

XIII Supernovae in binary stars 186 

XIV Black holes from supernovae 195 
XV Final chapter 206 

List of tables 

I The historical supernovae (p. 14) 

II Some supernova remnants (p. 123) 

III Magnitudes of Tycho's supernova (p. 136) 

IV Magnitudes of Kepler's supernova (p. 137) 
V Creation of the elements (p. 176) 

VI X-ray binary stars (p. 194) 

VII Velocity of escape (p. 204) 

I Supernovae in space and time 



»rom time to time bright new stars called supernovae 
flare briefly in the sky. In fact, there have been only five supernovae seen 
by man's unaided eye in the last iooo years. But supernovae have an 
importance in astronomy which transcends their mere numbers. Why ? 

The Universe is sending us at this moment virtually all the information 
that it ever will send us. Anything that we can find out about the 
Universe we can probably find out now. But understanding does not 
come easily. The Universe does not arrange itself to make itself com- 
prehensible. All the pieces of all the cosmic jigsaws are there, but 
scrambled and mixed up with scraps which have no significant place in 
that picture. Supernovae are a large and significant piece of the cosmic 

But even fundamental clues to the structure of the Universe can be 
overlooked, unrecognized when seen by an uncomprehending eye. For 
example the appearance for many months in ad 1006 of a supernova 
which shone so brightly that it cast shadows on the ground apparently 
did not change the beliefs of those who thought that the heavens were 
immutable. However two similar bright supernovae in 1572 and 1604 
threw light into the corners of minds ready to understand that the stars 
were not permanent, and the new astronomy began. 
At first astronomers concentrated on determining only the motions of 


the stars and gave fleeting attention to determining their life cycle. This 
was because although we can talk of stars having birth, life and death, 
they go through these stages on a very long time scale and changes are 
not usually apparent. The Sun has been as it is for some four billion 
years. Over a man's lifetime, indeed, over the lifetime of Man, most 
stars have remained very much the same. But recognizing from the 
occasional appearance of so-called new stars that the heavens do change, 
astronomers have sought to understand how stars evolve. It has not been 

Astronomers see very many different kinds of stars in the sky ; their 
objective in looking at these stars is first to explain how each kind works 
and what gives it the appearance that it has to us, and second to see how 
it changes from one stage to another. 

An astronomer has been compared with a Martian presented with a 
snapshot of a forest. The Martian must try to understand from his 
photograph the life cycle of a tree. He needs to know how an acorn turns 
into an oak sapling, and how an oak sapling turns into a fully mature 
tree; how the tree dies, falls and decays on the forest floor. Without 
actually seeing the processes of change in a tree, the Martian must study 
the different organisms that he finds and try to classify them into their 
different categories, guessing how one changes to another. 

In the same way, the astronomer studies the various types of stars, 
tries to classify them under different headings, and tries to explain how 
one type of star changes to another. Remarkably, after only ioo years of 
studying the intrinsic nature of stars, astronomers now believe that they 
do have a workable system for explaining how one kind of star changes 
into another. They believe that within a general outline, they have 
explained the whole lifetime of typical stars : from how they are born 
and begin to shine, up to the moment when all their energy is used up 
and they cease shining brightly and die. 

It is only very recently that astronomers have begun to study the 
death of stars. Unlike the majority of a star's life, which is long and 
peaceful, the stage which marks the death of a star can be brief and 


dramatic— so brief that it may last just a few months and so dramatic 
that for a few days the star outshines all its millions of neighbors put 
together. The lifetime of stars is so long and this death so brief that on 
average just one such event in a galaxy of i oo billion stars can be witnessed 
in the lifetime of a man. 

The bright explosion which marks the death of a star is a supernova. 
Where no star was seen before, astronomers see a bright new star 
shining. When first recognized such objects were named novae (pro- 
nounced "novee") meaning "new stars," although astronomers now 
understand that the star is not new but was so faint that it was not noticed 

In 1935 it was formally recognized that there were at least two very 
distinct kinds of nova, one much brighter than the other. The fainter 
kind is probably caused by a relatively weak explosion on one of a pair 
of stars orbiting each other, just as the Moon orbits the Earth. The 
brighter kind (which is 100,000 times brighter) is a supernova, and it is 
this explosion which marks a stellar death. 

In our own Galaxy of 100 billion stars only five supernovae have been 
witnessed in the last 1000 years, none since the invention of the telescope 
in 1608. But astronomers have found in the sky the remains not only of 
these supernovae but of others which occurred many thousands of years 
ago. A supernova explosion produces two visible kinds of objects. At the 
site where a supernova occurred astronomers see the shell of the 
exploding star speeding into space in fragments, colliding with tenuous 
gas in space and glowing from the force of the collision. This is called 
a supernova remnant. At the center of a supernova remnant may sit the 
hard core of the star that died, a star so faint that most emit no discerned 
light, a star packed so tightly by the force of the explosion that a match- 
box full weighs a billion tons. It is called a neutron star. 

The most-studied example of a supernova and its remnant is known 
as the Crab. Seen as a bright star in ad 1054, the Crab supernova 
produced a nebula which was discovered in the 18th century. At the 
center of the Crab Nebula lies a faint star, the neutron star produced 


by the supernova. The star is spinning on its axis at a rate of 30 revolu- 
tions per second, just as the Earth rotates once per day. A "hot spot" 
on the neutron star shines like a lighthouse into space and, as the beam 
passes across Earth once each revolution of the neutron star, it is per- 
ceived to flash or pulse. It is a pulsar. 

The study of supernovae has shed light on other unexplained 
problems in astronomy, such as how the elements came to be formed, 
including those in our bodies, and on the origin of cosmic rays, which 
are speeding particles of matter in space. 

Astronomers believe that supernovae are at the origin of cosmic rays 
and thus at the origin of part of the natural level of radioactivity on 
Earth. Some even speculate that past supernovae have played a part, 
through increasing radioactivity due to the cosmic rays, in the evolution 
of life itself. 

Thus supernovae, worth studying in their own right, have wide- 
ranging links with other studies as well as occupying a central position in 
the science of astronomy. Astronomers find it worthwhile to spend time 
studying not only supernovae in other galaxies, and the remnants of 
supernovae which have occurred in the Galaxy, but to delve into the 
tantalizing historical records of past galactic supernovae, attempting to 
discover the galactic supernovae which caused the remnants they see. 

Research into supernovae is only partly a matter of library study. 
Most information comes from investigating the sky. Practically every- 
thing has been achieved that can be achieved just by looking, and since 
the Universe is sending us all the information it ever will, it is only by 
new techniques that astronomers can achieve any new understanding. 
Some of the advances come from building bigger and better telescopes 
to perceive the light from fainter stars with finer detail. But the optical 
astronomer looks at the Universe through a restricted window in the 
atmosphere of the Earth. He does not see cosmic ultraviolet light, since 
it is absorbed by ozone in the atmosphere, and he does not easily see 
cosmic infrared light since this is absorbed by water vapor and oxygen. 
Just before World War II a second window on the Universe was 


opened: the window penetrated by radio telescopes. The brightest 
"star" seen by radio astronomers turned out to be a supernova remnant 
formed just 300 years ago by an unseen supernova. Among the other 
bright radio "stars" is Taurus A, the radio astronomer's name for the 
Crab Nebula, remnant of the supernova of 1054. Among the fainter 
radio stars were discovered the pulsars, now known to be neutron stars 
formed in supernova explosions. 

No further wide windows onto the Universe are available to the 
Earth-bound observer. To see the Universe of stars through other 
windows the astronomer flies his telescopes above the atmosphere. The 
first X-ray "star" to be identified, seen by a rocket-borne X-ray tele- 
scope, was the Crab Nebula. But it has been the previously unknown 
X-ray stars which have been the greatest surprise: the so-called 
compact X-ray stars are neutron stars, such as the one in the Crab, and 
more bizarre, the black holes, also formed in supernova explosions. 

The contemporary study of astronomy has been described by Geoffrey 
Burbidge as being divided into the study of the Crab Nebula and the 
study of everything else. An exaggeration of course, but this remark 
spotlights the central place held by the Crab and other supernovae in 
the new astronomy. This book begins with the search for the historical 
supernovae, and especially the progenitor of the Crab Nebula, the 
supernova of 1054. 

II Guest stars: 

the historical supernovae 


.he guest star of 1054 

In the first year of the period Chih-ho, the fifth moon, the day chi-cli'ou, a 
guest star appeared approximately several inches southeast of T'ien-kuan. 
After more than a year it gradually became invisible. 

In these straightforward words Toktaga and Ouyang Hsuan, the 14th 
century Chinese authors of the official history of the Sung dynasty, the 
Sung Shih, noted the appearance of a previously unknown bright 
star in the constellation now known as Taurus, the Bull. The day referred 
to is what we would now call July 4, 1054, and the star T'ien-kuan is 
what present-day astronomers call Zeta Tauri. These prosaic details pin 
down the precise day on which occurred an astronomical event whose 
effects are still with us over 900 years later. 

To the Chinese, guest stars were well worth noting, and indeed 
looking out for. They believed that man lived on Earth in a kingdom 
roofed with stars, and that man's destiny was subject to a "cosmic 
wind." Chinese emperors appointed court astrologers who watched the 
sky to ascertain the direction in which this cosmic wind would blow 
their subjects. These astrologers had been noting down celestial events 
since the 14th century bc, as these were believed to mark events of great 
significance in earthly affairs such as the death of princes. 


The observations were not noted down as incidental asides but as part 
of a deliberate policy. According to the Jesuit Lecompte's account in 
ad 1696 of the Ch'ing astronomical bureau: "They still continue their 
observations. Five mathematicians spend every night on the tower 
watching what passes overhead. One gazes towards the zenith, another 
to the east, a third to the west, the fourth turns his eyes southwards and 
a fifth northwards, that nothing of what happens in the four corners of 
the world may escape their diligent observation. They take notice of the 
winds, the rain, the air, of unusual phenomena such as eclipses, the 
conjunction or opposition of planets, fires, meteors and all that may be 
useful. This they keep a strict account of, which they bring in every 
morning to the Surveyor of Mathematics, to be registered in his office." 

The medieval Chinese historian Chang Te-hsiang writes in the 
Sung hui-yau that soon after the guest star of 1054 became visible, the 
Directoi of the Astronomical Bureau, Yang Wei-te, presented himself 
prostrate and kow-towing before his Emperor to tell him of its appear- 
ance. Perhaps Yang was fearful of not having foretold the coming of the 


Chinese oracle bones were made from an animal's shoulder blade and inscribed 
with a question. After searing the bone with a red hot poker, the answer to the 
question was divined from the pattern of cracks which appeared. The appearance 
of a guest star might have been held to confirm the answer. This bone dates from 
1300 BC and reads: " On the fth day of the month a great new star appeared in 
company with Antares." It is the first record of a "guest star" which could have 
been a supernova. 


guest star, since he assured the Emperor that because the star did not 
conflict with the constellation Pi (the nearby Hyades star cluster) and 
was bright and lustrous, it meant that a person of great wisdom and 
virtue was to be found in that part of China. This oblique compliment 
was no doubt well received by the Emperor and his assembled court, 

The supernova 0/1054 occurred in the Milky Way near the 
Chinese constellation Thien-Kuan, north of S hen, the present 
constellation Orion. It " did not conflict with Pi " {the Hyades) 
noted one astrologer, reassuringly. 


guest stars: the historical supernovae 

which acclaimed Yang's remarkably accurate prognostication, for Yang 
requested that it be filed at the Bureau of Historiography, perhaps the 
better to be retrieved if he should displease the Emperor at a later date 
and have to re-convince him of his loyalty. 
The account of this episode goes on : 

During the third month in the first year of the Chia-yu reign period [March 
ig to April ij, 1056] the Director of the Astronomical Bureau reported 
"The guest star has become invisible, which is an omen of a guest 's depar- 
ture." Originally, during the fifth month in the first year of the Chih-ho 
reign period, the guest star appeared in the morning in the east, guarding 
T'ien-kuan. It was visible in the day, like Venus, with pointed rays in 
every direction. The color was reddish-white. It was seen like that for 
twenty-three days altogether. 

The guest star was also seen and recorded by Japanese astronomers. 
They noted its appearance, as bright as the planet Jupiter, in early 
June 1054. It therefore seems that they may have observed it before it 
was at its brightest (as bright as the planet Venus) which occurred early 
in July 1054 when it was discovered by Yang Wei-te. Possibly this is 
why he seemed somewhat apprehensive when making his report to the 
Emperor. He may well have had a real danger to fear — according to 
legend, the Chinese astronomers Ho and Hi were beheaded for failing 
to predict the solar eclipse of 2137 bc. 

The star in an Arizona cave 

A rather strange thing about the new star of ad 1054 is tnat 
no written records of it can be found in the writings of other cultures. 
European, Arabic and Korean chronicles remain silent about an event 
which must surely have been an astronomical spectacle, far outshining 
such pale rivals as Halley's Comet, which appeared 12 years after. 
But in North America, despite the lack of writing at that time, there 


is a strong possibility that the new star actually was seen and recorded. 
The evidence was uncovered in the early 1950s by two astronomers on a 
non-astronomical expedition to rescue cultural relics in northern 
Arizona before the completion of a dam flooded a valley of the Colorado 

The astronomers were Helmut Abt, then at Yerkes Observatory, 
Wisconsin, and Bill Miller, who was chief photographer at Mt. Palomar 
Observatory, California. On their expedition they found two remarkable 
prehistoric drawings. Later on, similar drawings were discovered at 
other sites. 

The first drawing, in the White Mesa, was found on the wall of a cave 
and shows a crescent Moon with a circle, apparently representing a 
bright star, overlapping the lower cusp or horn. In the second drawing 
the Moon is also shown as a crescent (but reversed from the first) and 
the circular object is directly under the lower cusp. This drawing was 
found on a canyon wall near ruins in Navajo Canyon. A third shows a 
series of circles below the crescent possibly representing motion of the 
Moon with respect to the bright star. 

Because the first drawings were found in association with Pueblo 
Indian dwellings it is possible to give some indication of their date by 
archaeological methods. Pueblo Indian pottery from different locations 
and of different dates is highly distinctive. Potsherds collected at the 
two sites have been analyzed by Robert C. Euler, Curator of Anthro- 
pology at the Museum of Northern Arizona, who found that most of the 
inhabitants of the White Mesa site lived there later than ad 1070, but 
that there were earlier sherds in the collection. At the Navajo Canyon 
site a deep arroyo, or gulley, cut into the floor of the canyon, had exposed 
broken pieces of pot, with the oldest at the deeper levels. Euler deter- 
mined that the site was occupied by Pueblo Indians from before 
ad 700 to after ad 1300 with about a fifth of the sherds collected dating 
from between ad 900 and 1100. 

It is possible that the drawings represent nothing rarer than a near 
approach of the crescent Moon to one of the brighter planets, Venus or 


Contemporary views of the 1054 supernova? Two ancient Indian markings 
discovered by William C. Miller. He found the one at left at the White Mesa of 
Arizona in 1952 and the one at the right at Navajo Canyon in 1954. 

Jupiter perhaps. Such a so-called conjunction between the Moon and a 
bright planet takes place several times a year; it is an event which has 
for a long time attracted the attention of astrologers from many cultures 
because the Moon and planets are supposed to exert influences which 
augment or conflict at times of conjunction. The crescent Moon and 
stars is therefore a common symbol; there are more than a dozen 
representations on present day national flags— Turkey's, for example. 

There would have been many times when such a conjunction occurred 
during the centuries in which the Arizona sites were occupied. If this 
recurrent event was of interest, however, it is strange that just a few 
pictographs have been found in Pueblo Indian sites. On the other hand, 
the scale of the circle relative to the Moon in both drawings suggests 
that the star was comparable in brightness to the crescent Moon, and 
certainly the guest star of 1054 was one of the most notable stars known 
to have appeared during the time the sites were occupied. Cambridge 
astronomer Fred Hoyle suggested this to Bill Miller, who then calcu- 



lated the position of the Moon during July and August 1054. He found 
that on July 5, 1054 at 3 o'clock in the morning (that is, the day after 
the guest star was noted at its brightest by Chinese astronomers) the 
crescent Moon would have appeared just above the guest star in the 
morning twilight, and could have been seen from the Pueblo Indian 
homes, which had an unobstructed view to the east. 

The evidence that the drawings are representations of the new star of 
1054 is circumstantial and there are some difficulties — why, for instance, 
is the crescent Moon reversed in one picture? (Miller suggests that 
non-astronomically trained people do not worry about this kind of thing 
and remarks how common it is to find the Moon shown in the wrong 
orientation in modern illustrations.) But it may well be that these 
drawings really were made by Pueblo Indians 438 years before Columbus 
sighted land in the Americas, and are indeed records of what amounts 
to a remarkable pre-Columbian Independence Night fireworks display. 

What can be learned about the nature of the guest star from the 
ancient observations ? We have first of all its approximate position near 
to the star Zeta Tauri.* But more than this, the Sung Shih, quoted at the 
beginning of the chapter, says in another passage about the guest star 
that it "remained" in the sky near Zeta Tauri for almost two years. 

The ancients recognized two kinds of stars, the fixed stars and the 
wandering stars, or planets. It is now known that in fact both the fixed 
stars and the planets are moving through space at speeds comparable 
with one another. But the wandering stars — planets — appear to move 
faster because they are much closer to us than the so-called fixed stars. 
The planets, in fact, belong to the nearby solar system, whereas even 
the closest star is 10,000 times more distant than the farthest planet. 

Similarly, comets too speed through the solar system and wheel 
across the sky, typically in a few months. So even though the Chinese 

* The present system of naming bright stars uses the Latin constella- 
tion name and a Greek letter, with Alpha (a) usually denoting the 
brightest, Beta (£) the second brightest and so on. Zeta Tauri is there- 
fore the sixth brightest star in the constellation of Taurus, the Bull. 


guest stars: the historical supernovae 

called some comets "guest stars" and did not exclusively use the term 
for new fixed stars, the fact that the guest star was noted to be stationary 
for two years places it certainly as far as the edges of the solar system, 
beyond the most distant planet, Pluto, and probably outside the solar 
system altogether. 

The Chinese and Japanese records also tell us something of the change 
in brightness of the guest star. We can deduce from the extracts quoted 
above that a month before the maximum brilliance of the guest star it 
was as bright as Jupiter; that at its brightest it rivaled Venus, visible 
during daylight; that the guest star ceased to be visible in daylight 
23 days after that; and that it was finally fading from the evening sky 
630 days later. 

Because the guest star faded from view to the naked eye into the 
evening twilight in 1056, it probably was not as faint as the faintest stars 
visible on the darkest nights, but somewhat brighter. 

The evidence that we have of the length of time the star was seen is 
all consistent with describing this guest star as a supernova, rather than 
as a nova, since novae fade more quickly. 

It is now known as the supernova of 1054. It was one of only five 
supernovae which can have been seen by the unaided eye of man within 
the last thousand years. 

Just as the guest star faded from view, so its memory faded from 
men's minds. But the vital details of its visibility had been recorded, were 
dutifully stored and recopied by Chinese scribes over the centuries, and 
can now help in the attempt to understand the mysteries of the Crab 
Nebula, left over from the original momentous explosion. 

Sweeping, bushy stars eliminated 

What other supernovae were recorded among the oriental 
observations? To distinguish records of supernovae from records of 
other celestial phenomena such as comets, aurorae, meteors, lightning 
and so on, astronomer David H. Clark and historian F. Richard 



Stephenson have united in an interdisciplinary study. They have iso- 
lated all mentions of guest stars which had no motion and so excluded 
moving objects such as comets and meteors. Many objects could be 
eliminated as being described as "bushy stars" or like a broom ("sweep- 
ing stars"), references to a comet's tail (though the records are not 
always unambiguous on this point). This gave them a list of 75 probable 
novae or supernovae sighted between 532 bc and ad 1604. Among these 
they identify six supernovae which occurred before ad 1500 (Table I). 

table I The historical supernovae 










20 months 


msh 14 —63 



3 months 

GII.2 —O.3? 



8 months 

G348.7 +0.3? 



Several years 


PI459 -41 



24 months 





6 months 

+ i? 


Clark and Stephenson recognized them because they are noted in the 
records as lasting longer than an ordinary nova and also were seen along 
the central line of the Milky Way, which is where supernova remnants 
are found and hence the only place where supernovae occur. Apart from 
the guest star of 1054 there are two supernovae, recorded in oriental 
observations, whose history is known in sufficient detail to have enabled 
radio astronomers to claim with some confidence that they have identi- 
fied their remnants. They are the supernovae of 1006 and 1181. 

The supernova of 1006 

Probably the brightest star to have appeared in the sky for 
the last thousand years was the supernova of 1006 which blazed from 

guest stars: the historical supernovae 

the southern constellation of Lupus. Because it was so far south and 
was below the horizon to northern European observers, our main 
sources of information are Arabic, Japanese and Chinese texts. These 
texts agree fairly well on the position of the phenomenon but it can 
readily be admitted that they are obscure over other details of its 
appearance. In spite of the difficulties in interpreting and collating the 
ancient sources, modern astronomers are convinced that the star was a 

The main eye witness source is the Egyptian Ali b. Ridwan (who died 
in 1061), who lived in the old city of Cairo. He mentions the supernova 
in an autobiographical footnote to an astrological work by Ptolemy, the 
Tetrabiblos, which he was editing. He says that it appeared "at the 
beginning of my education" in the 15th degree of Scorpio. The word he 
uses for the star is nayzak which is fairly rare in Arabian astronomical 
works. Where it does occur, it refers to a very bright comet, which is 
how later commentators translated the word. Ali, however, says that the 
object remained stationary relative to the other stars, while the sun 
moved into Virgo. Therefore, it was not a comet. 

Ali included a list of the exact positions of the planets at the time of 
his first sighting the supernova. From his list we can be quite sure of the 
date: April 30, 1006. The modern computation of the planetary positions 
on this day agrees well with Ali's own data. Ali also includes the expected 
astrological conclusions. In this case, famine, death and pestilence broke 
out. Ali detailed his observation of the supernova : 

It was a large nayzak, round in shape and its size two and a half or three 
times the size of Venus. Its light illuminated the horizon and it twinkled a 
great deal. It was a little more than a quarter of the brightness of the Moon. 

It is not clear what Ali means when he says that the supernova was 
about three times the "size" of Venus. Astronomers then thought that 
the brighter stars had perceptible disks, this being a physiological effect 
in the eye. Two Chinese sources even say that it was a half Moon. 



Ibn al-Jawzi, another Arab source but from the 13th century, says that 
"it was a large star similar to Venus." It seems probable that these 
astronomers, in comparing the supernova to Venus, were trying to give 
an impression of its brightness compared with the brightest star they 
knew. Ali, in fact goes further and says it was a quarter of the brightness 
of the Moon. This may mean a quarter of the brightness of the Moon 
when only a quarter is illuminated. This latter interpretation is more 
consistent with the comparison to the size of Venus. 

Ali and Ibn offer the seemingly independent observations that the 
light from the star "illuminated the horizon" and that "its rays on the 
earth were like the rays of the Moon." A Chinese observer noted in 
the Sung Shih that it "cast shadows." Another Chinese source says that 
"it shone so brightly that objects could be seen by its light." This all 
indicates that it was much brighter than Venus. 

The most significant European account is from Hepidannus, a monk 
of St. Gall in Switzerland. In his Latin Chronicle of 1006 he says that 
he saw the star in the extreme south. This places a limit on how far 
south it could have been since it must have been above the horizon as 
seen from St. Gall (latitude 47^ degrees). Chinese and Japanese 
observers place it on the present day border between the constellations 
Lupus and Centaurus. Its position can thus be tied down to within fine 

The color is doubtful. One Chinese observer, the Director of the 
Astronomical Bureau at the Imperial Court of the Emperor Chen-tung, 
called it "a large star, yellow in color." This is not to be relied on, 
however, as an objective description. The reason is partly astrological 
and partly political. When the supernova appeared, it was so striking 
that everyone in the Chinese capital, Kaifong, was filled with alarm, and 
the general opinion was that it was a very bad omen which would be 
followed by famine and plague. The Director of the Bureau of 
Astronomy, Chou K'o-ming, was out of town at the time. When he 
returned, he found the Emperor very anxious and distressed by the 
situation. Having considered the evidence, he announced that the star 


GUEST stars: the historical supernovae 

belonged to the astronomical category Chou~po. This was an excellent 
move as such a star was an omen of prosperity and could occur only in 
the reign of a wise and just monarch. The Director was later promoted 
to Librarian and Escort of the Crown Prince. The important characteris- 
tic of a Chou-po star from our point of view, however, is that it was 
always yellow in color. The description is therefore inevitable, given the 
classification, and does not provide us with scientific information. A 
Japanese report that it seemed to be blue-white may be more objective. 
When we ask how long the supernova was visible, we find difficulties, 
though it was long-lasting. The Chinese Chu-Su says that it "later 
increased in brightness" but does not say later than what. Ali, however, 
says that it disappeared suddenly. We are told by several writers that the 
supernova was visible for three and a half months after which it was too 
close to the Sun to be seen, but that would cause a gradual disappearance, 
so probably Ali is not referring to this. Venus in the same position would 
be clearly visible in daylight; therefore we can say that the supernova at 
that point could not have been brighter than Venus. 

After seven months behind the Sun's glare, the supernova reappeared 
in the dawn sky between November 24 and December 22. How long 
after this the star remained visible is difficult to determine, but it seems 
to have been more than a year. The Chinese chronicle, the Sung Shih 
refers to "a Chou-po star" in November 1006 and again on May 15, 
1016. No positions are given but if both stars are the same one it must 
have been erratically visible for up to 10 years. The only indication that 
both the references are to the 1006 supernova is that these are the only 
instances of "Chou-po" being used to describe a guest star. The Sung 
Shih gives more evidence that the supernova was visible for several 
years. In a passage specifically describing the supernova of May 1006, 
it mentions the first disappearance and reappearance and continues with 
the significant word "thereafter" to say that it disappeared near the Sun 
in the eighth month and reappeared in the eleventh month. Because this 
suggests a continuing phenomenon, being hidden by the Sun annually, 
the star must have been visible for at least two years. 



We may be trying to give too precise a description from those frag- 
mented notes recopied by generations of scribes and summarized by 
medieval historians from the bits and pieces salvaged after the Mongol 
invasion of China in 1345. Nevertheless, the location is relatively clear 
and has enabled two radio astronomers, Frank Gardner and Doug 
Milne, to identify in 1965 a radio source at the area where the 1006 
supernova appeared. Using the 210 foot radio telescope at Parkes in 
New South Wales, Australia, they found a structure closely resembling 
other supernova remnants. 

In 1957 Walter Baade had tried to find visible traces of gas where the 
supernova had appeared, just as he had previously found the nebulae left 
by other supernovae, but was not able to make any positive identifica- 
tion. The object was too far south for the Californian telescopes. In 1976, 
however, Sidney Van den Bergh, using the Cerro Tololo telescope in 
Chile where the constellation of Lupus passes overhead, discovered a 
faint wispy nebula close to the radio source, ejected from the supernova 
when Ali b. Ridwan was a student. 

The supernova of 1 181 

Another supernova flared up and briefly amazed observers 
in ad 1 1 81. The evidence is found in Chinese and Japanese chronicles 
and is difficult to interpret. Finding out as much as possible is worth 
some effort, as radio astronomers have found a supernova remnant 
called 3c 58 in a position which the records indicate was the position of 
the 1 181 supernova. We must fit together, as well as we can, the observa- 
tions of the ancient astrologers and astronomers with the work of the 
20th century radio astronomers, in order to form a picture of the violent 
death of the star and its subsequent gradual dissipation into the inter- 
stellar matter. 

The records present us with the two most common difficulties of this 
sort of work. One is that they do not agree. The other is that what they 
do say is imprecise. The star was observed from three geographical areas 
and we can make some deductions from the three sets of records. 

guest stars: the historical supernovae 

This radiophotograph ofjC^8, the remnant of the supernova of AD 1181, has 
been made from observations with the Westerbork Synthesis Radio Telescope by 
A. S. Wilson and K. W. Weiler. Like the Crab Nebula it has aflat appearance, 
rather than the shell-like look of most other supernova remnants. Four times more 
distant than the Crab, it is almost six times its size although of comparable age, 
and so has exploded at six times the average speed of the Crab. No optical 
nebula has been found at this site. 

The first to see it were the observers in southern China. The Sung 
Shih tells us that the star was first seen on August 6, 1181. The next to 
see it seem to have been the Japanese. The History of Great Japan, 
written in 1715, says that a guest star appeared in the north on August 7, 
1 181. The Chinese in the northern Chin empire reported in the History 
of the Chin Dynasty seeing the star on August 11. 

How long the star was visible is of critical importance in deciding 



whether or not it could have been a supernova identifiable with the 
known radio source. The Sung Shih says that the star was visible until 
February 6, 1182 "altogether 185 days; only then was it extinguished." 
The account in the Chin Shih gives a somewhat shorter time, 156 days. 
The duration of several months makes it a reasonable assumption that 
the star was a supernova. 

The three groups of accounts state that the supernova appeared in or 
near different Chinese constellations. However, all lie within the 
present-day constellation of Cassiopeia. Chinese constellations were not 
firmly defined and the discrepancies are not significant, just disappoint- 
ingly vague. 

The evidence for the color and brightness of this star is largely 
Japanese. The Azuma Kagami tells us : 

At the hour hsu [19-21 hours local time] a guest star mas seen in the north- 
east. It was like Saturn and its color was bluish-red and it had rays. There 
had been no other example since the third year of Kanko [the supernova of 

AD IO06] 

The comparison with Saturn is striking because Saturn would not 
actually have been visible between those hours, but only towards 
dawn. Mars would have been visible at the time, and therefore the 
mention of Saturn seems to deliberately imply that the supernova's 
brightness was closely comparable. We do not know, however, whether 
this was the supernova's maximum brightness or whether it was 
discovered before maximum. Furthermore, this Japanese chronicle says 
that there had been no other star of its kind since 1006. The supernova 
of 1006, as we know, was very bright. Perhaps the supernova of 1 181 
became very bright too. This interpretation would be consistent with 
the considerable interest which the star aroused and the number of 
references to it. But the evidence is not clear enough, and presumably 
never will be. 


The Star of Bethlehem 

The best known bright star in history is the Star of 
Bethlehem. Was it a supernova ? To find out, we have to examine the 
documentary evidence in the same way as we have done for the Chinese 
and Arabic accounts of other supernovae. 

At the birth of Jesus, according to the gospel of St. Matthew, chapter 
2, "there came wise men from the east to Jerusalem saying, where is he 
that is born King of the Jews ? For we have seen his star in the east and 
are come to worship him." The term translated here as "in the east" 
means more precisely "at its heliacal rising," that is, the wise men saw 
the star appear in the first rays of dawn. 

From Jerusalem the Magi, who may have been astrologers from 
Persia or from the Tigris-Euphrates valley civilizations of Assyria, 
Mesopotamia, or Babylonia, traveled south to Bethlehem following the 
star which "went before them, till it came and stood over where the 
young child was." This passage is difficult to identify with any astro- 
nomical phenomenon since the motion of stars is generally east to west 
and astronomical objects are so distant that they do not identify one 
particular terrestrial location but stand equally over whole areas when 
at the zenith. 

Setting this aside for a moment, however, the third item of evidence 
is from the Protoevangelium of James (21:2), one of the Apocryphal 
gospels not included in the Bible, which offers the following: 

And he [Herod] questioned the wise men and said to them: " What sign did 
you see concerning the new-born King?" And the wise men said: " We saw 
how an indescribably great star shone among these stars and dimmed them, 
so they no longer shone, and so we knew that a King was born for Israel." 

Though specific as to the brightness of the star, which would be 
comparable to the brightness of the Full Moon if it flooded the sky with 
its light and rendered surrounding stars invisible, this passage raises the 
difficulty as to how Herod and his advisers could have come to miss 



noticing the star, unless Herod's question is deliberately disingenuous. 

When did the birth of Jesus occur ? The first naive attempt at an 
answer is December 25 in the first year of the Christian era. However, 
the tradition that Jesus was born at midwinter began about ad 336, 
possibly because Christians wished to hide their celebrations among the 
general festival of Saturnalia, or, more likely, because the Church drew 
the existing pagan festival within the Christian tradition. Luke (2:8) 
says that at the time of Jesus' birth, shepherds were " abiding in the 
fields, keeping watch over their flock by night." During winter in Judea 
flocks were penned, being set free in the spring and guarded by night 
in the lambing season (March and April). 

As to the year, the presently accepted calendar of years is from a 
correlation between Christian tradition and Roman imperial history 
which is expressed in a calendar reckoned ab urbe condita (auc), from 
the founding of the city of Rome. The correlation which has been 
adopted is by Dionysius Exiguus (ad 525) who missed the year zero 
between 1 BC and AD 1 and forgot the four years during which Emperor 
Augustus ruled under his own name of Octavian. This puts Jesus' birth 
in 5 bc or auc 749. Herod died just before passover in 4 bc. Jesus was 
therefore certainly born before then. In his account of Jesus' birth, Luke 
says that Caesar Augustus had ordered a tax and that this was why 
Jesus' parents had to travel to Bethlehem. Such an order was issued in 

8 bc; the tax would have been collected in the years following. Luke says 
that tax collection was begun when Quirinius was governor of Syria, but 
he was not governor until ad 6, although he was an Emperor's legate in 
Syria between 6 and 5 bc, and Luke may have been confused as to his 
rank. Tertullian, a Roman historian, says that the census at the time of 
the birth of Jesus was taken by Saturninus, who governed Syria between 

9 and 6 bc. 

It seems that the birth of Jesus occurred one springtime between 7 
and 5 bc. This excludes Halley's comet as the star (it appeared in the 
autumn of 12 bc). Another significant astronomical event was a conjunc- 
tion of Saturn and Jupiter in the constellation Pisces in 7 bc. First 


guest stars: the historical supernovae 

suggested as the possible Star of Bethlehem by Kepler, this was a similar 
conjunction to the one which led to the discovery of Kepler's supernova. 

The possibility that the Star of Bethlehem was a supernova (or nova) 
is one which occurred to astronomers including Tycho Brahe after the 
appearance of the supernova of 1572, which was interpreted by some as 
signifying a further event of the same kind, possibly the second coming 
of Christ. What was possibly a nova or supernova was recorded in the 
History of the Former Han Dynasty as occurring in late March or early 
April in 5 bc, and lasting over 70 days. It appeared in what is now 
called the constellation of Capricorn. In springtime this constellation 
rises some five hours before the Sun so that the star would have been 
first observed in the rays of the dawn, as Matthew implies. The rela- 
tively short length of time for which the star was visible suggests that it 
was a nova rather than a supernova. The Chinese records actually call 
the object a "broom-star" which is usually used for comets having tails 
like a brush, although, as pointed out by D. H. Clark, J. Parkinson and 
F. Stephenson, one record of the well known supernova of 1572 mis- 
classifies this star in the same way. Whether nova or comet, it was 
probably not a supernova. 

The notion that the Star of Bethlehem was a nova or supernova 
survives in modern literature in a story {Nova) by Arthur C. Clarke. In 
Clarke's story, a Jesuit astronomer/astronaut finds his faith severely 
shaken by the discovery of the archaeological remains of a beautiful, 
peaceful and cultured civilization, exterminated when their sun exploded 
in a supernova explosion. His astronomical training leads him to calcu- 
late the date of the catastrophe, only to find that it was apparently timed 
by God to occur so that the light of the explosion reached Earth at just 
the right moment to proclaim the birth of His Son. 

Apart from fiction, contemporary chronicles of the Chinese or other 
peoples might have been expected to record the supernova, if there was 
one, especially if it was as bright as the Full Moon. No such records have 
been found. Furthermore, a supernova would not have had the unusual 
motion attributed to it in Matthew 2:9. 



It is of course possible that the Star of Bethlehem was a genuinely 
miraculous event for which there is no physical explanation, in which 
case the above analysis is in vain. It is also possible that the narrative is 
what in Jewish tradition is called a midrash, a historical equivalent of the 
Christian sermonizing, that is, a historical fact presented in a popular 
manner with decorations adapted to the reader's expected mentality, 
with echoes of previous parallel events such as, in this case, the birth of 
Moses or Abraham. British astronomer David Hughes writes "no king 
worth his salt in those days was born without some celestial manifesta- 
tion. A star greeted the birth of Mithridates (131-63 bc) and Alexander 
Severus." This is probably the way to reconcile the difficulties in the 
various records, which do not have the physical self-consistency of, say, 
the Chinese accounts of the supernova of 1054, and cannot be taken to 
compel the same conclusion. 

The Star of Bethlehem might perhaps have been the nova of 5 bc but 
there is no evidence that it was a supernova. 

Since bright supernovae do not often explode within sight of the 
Earth, and a man is unlikely to see one in his life, the impact that they 
made on the ancient world may perhaps have been much greater than 
surviving records imply. In the Renaissance, however, two circumstances 
combined to make the modern astronomer look there with particular 
interest when on the track of supernovae. One was the appearance very 
close together of two spectacular supernovae. The second was a new 
interest in observing and recording natural phenomena for their own 
sake. For these reasons, the Renaissance supernovae were of immense 
significance both at the time and to us now. 


Ill The Renaissance supernovae: 
shattering the crystal spheres 

■ x,. 

^spite the evidence of their own eyes, most people 
today believe what astronomers tell them about the way the world 
moves. From an early age we are all taught that the Earth is round, not 
flat, and that like the other planets it spins on its own axis and orbits the 
Sun. Yet every day we see the Sun move through the sky, along with the 
stars and planets. Few have time to prove to themselves that the Earth 
really does move round the Sun. 

Four hundred years ago, astronomers assured those who listened that 
the Earth was stationary, and orbited by the Sun. Our ancestors just 
20 generations away believed this. Supernovae helped change their 

Crystal spheres and epicycles 

For 2000 years, most Europeans accepted the description 
of the Universe that had been given by the Greek philosopher Aristotle 
(381-322 bc). Ptolemy of Alexandria turned Aristotle's notion into a 
quite workable mathematical form in the 2nd century ad. Then in the 
13th century, Thomas Aquinas took Ptolemy's system as the correct 
picture of the Universe to give a background to his ideas on theology. 
When Aquinas' theology was accepted by the Roman Catholic 


1 ^W/tti^VJO^ 

/Vter Apian' s picture of the Universe, published in 1524, illustrates the 
Aristotelian conception of the Universe fused with Christian theology. Outside 
all is the Coelum Empirreum, the highest heaven, home of God and the Elect. 
Ten spheres are nested within, starting with the Primum Mobile, inside which 
are the ninth sphere of crystal, the eighth sphere containing the Firmament of 
fixed stars and the spheres of the seven planets, including number 4, the Sun. 
Below the sphere of the Moon are the terrestrial regions of fire, the cloudy sky 
and, at the center of all, the Earth itself. 



Church, the cosmologies of Aristotle and Ptolemy became woven into 
the social order which was seen as a chain stretching from God down to 
the humblest living creature. A king might hope to be secure on his 
divinely granted throne because order was seen in the heavens. The 
philosophy and the science were not questioned because if one group of 
certainties was shattered, all the rest, including the social order, would 
be subject to doubt. 

What was this cosmology which so constrained people's lives ? In the 
Ptolemaic system, the Universe has a set of spheres like an onion. The 
Earth was thought to be at the center of the visible Universe. Com- 
pletely surrounding the Universe, according to medieval philosophy, 
was the Empyrean where God sat enthroned, accompanied by the souls 
of the just. Within this lay the primum mobile, the First Mover. This was 
a sphere of an intellectual substance which was thought to be the cause 
of all movement in the heavens. Movement in the heavens was, in turn, 
the cause of all movement on Earth. 

Inside the primum mobile was the eighth sphere, or firmament, the 
sphere embedded with the fixed stars. (Some saw the necessity for a 
ninth sphere, the caelum igneum, or fiery sphere, to account for some of 
the apparent movements of the planets.) Below these outer spheres were 
the spheres of the seven planets, as Saturn, Jupiter, Mars, the Sun, 
Venus, Mercury and the Moon were then called. Each revolved at its 
own speed about the Earth, the spheres closest to Earth slipping behind 
the rotation of the firmament by the largest amount. 

Aristotle said that the sphere was the perfect geometrical form. This 
was why the heavenly bodies were embedded in spheres and why they 
logically had to move in circles. The lowest sphere contained the Moon 
and marked the boundary between celestial perfection and the sublunar 
region of death and corruption which mankind inhabited. With this 
picture, everything falls into place. Thomas Aquinas was able to link it 
with the Christian view of the Fall of Man and his propensity to sin. 

Even though the English philosopher Thomas Digges believed, like 
Copernicus, that the Sun was the center of the Universe, he still held to 



the theory that above the Moon all was unchangeable and that death and 
decay were possible only beneath, describing in 1576 

the Moon's Orb that environeth and containeth this dark star [Earth] and 
other mortal, changeable, corruptible Elements. 

John Donne (1571-1631) often referred to these heavenly spheres, 
even in love poetry : 

If, as in water stir'd more circles bee 
Produced by one, love such additions take, 
Those like so many spheares, but one heaven make, 
For they are all concentrique unto thee. 

But at the center of the spheres, wrote Donne, were "dull sublunary 
lovers" whose love could not survive absence because it was based on 
the senses alone and was therefore imperfect, in contrast to God's 
perfect love from heaven, above the highest celestial sphere. 

Given the preconception that above the Moon everything was perfect 
and eternal, the followers of Aristotle accounted for any changes which 
they saw occurring in the sky by insisting that they took place below the 
Moon. Comets, for example, were said to be atmospheric. According to 
Aristotle himself, comets were exhalations from the Earth produced by 
the burning of gases in the atmosphere above the Earth and set on fire 
by the Sun. Today's view, that they are celestial bodies traveling on 
elongated orbits which bring them in from the depths of space through 
the solar system, between the planets, would have been impossible in 
the Ptolemaic system. The comets would have had to penetrate the 
crystal spheres in which the planets were embedded, shattering the 
perfection of circular motions. In Aristotelian science, comets were in 
fact classed with rainbows,- gales, dew, lightning and all other atmos- 
pheric phenomena as being meteorological. It was a much later develop- 
ment which restricted the use of the word meteor to shooting stars. 



Although astronomical historian Owen Gingerich has cast doubt 
on this picture's historical authenticity, the medieval cosmology of 
the Universe in which the Earth was at the center of a sphere of 
stars is well illustrated in this alleged fifteenth century woodcut. 
A pilgrim looks through the Firmament at the mechanisms of the 
Primum Mobile beyond. 

Curiously, although oriental astronomers recorded dozens of novae, 
only two possible novae were known to European astronomers before 
1572. No systematic way had been developed to explain how such a 
temporary phenomenon could appear in the eternal firmament. 

Each of the two possible novae was explained away. The first was the 
Star of Bethlehem, but because it was thought to be genuinely miracu- 
lous it needed no general explanation. Secondly, the Greek historian 
Pliny records that Hipparchus is said to have observed a new star in 
134 bc. The usual explanation for this was that what he saw was 
actually a comet, although Hipparchus should have known the difference 
between the point-like appearance of a star and the diffuse appearance 


A pcrfit defcription of the CocIeftiallOrbes 

accenting t» the most Mncient dtZtrine of the PjtlsA- 
*ore*m. &c. 

** * * * 

+ * V ** 


Thomas Digges' Perfect Description of the Celestial Orbs was published in 
1576. While trying to come to terms with Copernicus'" idea that the Sun was at 
the center of the Universe, Digges was still a prisoner of the Aristotelian idea that 
the stars were immutable. The Orb of stars is described as " immovable . . . 
garnished with perpetual shining glorious lights innumerable" whereas the Orb 
of the Earth carries " this globe of mortality." 



of a comet (the word comet derives from coma, meaning hair). However, 
it was firmly held that whatever he saw, it was sublunary. 

So pleasing was Aristotle's description of the Universe, with its 
perfect circles and uniform speed of planetary motion, all within the 
constant firmament, that until the 16th century, and even after, most 
philosophers expended their energy in saving the appearances, that is, 
devising new methods of calculation and minor embellishments of 
Aristotelian principles which would enable the observations to be 
reconciled with the theory. 

Ptolemy in fact found that the movements of the planets which he 
observed did not fit the theory of exactly circular orbits. He elaborated 
Aristotle's theory by claiming that planets moved on epicycles, or small 
circles whose centers themselves moved in circular orbits. In a similar 
way, Ptolemy dealt with the problem that the observed speeds of the 
planets were not uniform at all points of their orbits, as according to 
Aristotle they should have been, since uniform motion was the perfection 
expected in a cosmic body. Ptolemy maintained that the speed of a 
planet was indeed constant, not as seen from the Earth, but as observed 
from another point in space, the equant. 

Later philosophers followed in Ptolemy's footsteps, devising elabora- 
tions of the epicyclic idea to "save the appearances." Even Copernicus 
who in 1543 took the drastic step of re-ordering the solar system so that 
the Sun, and not the Earth, was at the center of the Universe, still 
presented his new theory of the solar system in terms of epicyclic motion. 

By about this time it was becoming increasingly clear that the motions 
of the planets were getting harder to explain on a simple epicyclic 
system. More and more complexities had to be introduced in order to 
predict accurately the positions of the planets in the sky. But the 
ancient learning, supported by the Church, was deeply entrenched. 
Perhaps it is not surprising that philosophers were prepared to go to 
great lengths to preserve the old system, which to most people must 
have seemed the only logical one. 

Indeed, when Andrew Osiander came to write the preface to Coper- 



nicus' revolutionary book, he had to say that the new theory merely 
simplified calculations and did not necessarily mean that the Sun had 
ousted the Earth from the center of the Universe. But nonetheless the 
contradiction between theory and observation arose because of the 
Aristotelian presumption that the Universe above the Moon was perfect. 
It took the supernova which suddenly exploded in 1572 to shatter the 
crystal spheres. 

Observations of the supernova of 1572 

The first recorded observation was made on November 6, 
1572, by a Sicilian mathematician, Francesco Maurolyco, who observed 
a very bright, previously unknown star in the constellation of Cassiopeia. 
There was great excitement about the new star. " I am unable to admire 
enough the new shining of the star of our time" wrote Maurolyco. He 
noted that he saw the star at the third hour of the night and wrote down 
its approximate longitude and its angle above the horizon. There is 
some doubt about whether Maurolyco had seen the supernova before 
November 6, but a Spanish philosopher, Hieronymus Mugnoz, was 
teaching an outdoor class in astronomy on November 2 and said after- 
wards that he would certainly have noticed the new star in Cassiopeia if 
it had been visible then. The beginning of the supernova can therefore 
be placed between November 2 and 6, 1572. 

On November 7, Paul Heinzel of Augsburg, Bernhard Lindauer of 
Winterthur in Switzerland and Michael Mastlin of Tubingen, Kepler's 
teacher, also saw the supernova. Professor Mastlin satisfied himself that 
the nova was a star and not a comet. He did this by selecting two pairs 
of stars in Cassiopeia so that lines between the members of each pair 
would intersect at the supernova, which was accomplished by holding a 
thread before his eyes so that it passed through two of the known stars 
and the new star. In this way, he was able to say that the new star was 
not moving in relation to the other stars of Cassiopeia. Thomas Digges 
carried out the same experiment using a six-foot ruler, and placed the 


Tycho Brake *Sri 

star at the intersection of the line joining Beta Cephei to Gamma 
Cassiopeiae and Iota Cephei to Delta Cassiopeiae. 

Mastlin's and Digges' simple observations were elegant, but the man 
who won fame and fortune from the supernova of 1572 was Tycho 
Brahe. Brahe, a Danish astronomer, was an extraordinary man. His 
personal life was wild and undisciplined in the extreme. He had a gold 
and silver bridge to his nose necessitated by an injury suffered in a duel 
when he was a student. Later in his life he lost the private island 
observatory which had been granted to him by King Frederick because 
of his arrogant and unjust treatment of his tenants. As a scientist, 
however, he was entirely different. Far from being arrogant and bom- 
bastic, he was meticulous and precise. Indeed, his observations, made 
before the introduction of the telescope, are recognized as the finest ever 
made with the naked eye, and achieve an accuracy limited only by the 
acuity of the eye itself. 

Brahe was at the beginning of his career in 1572, and it was in fact the 
supernova which inspired him to devote his lifetime to making accurate 



measurements of the positions of the stars and planets. As Kepler, his 
pupil, said, "if that star did nothing else at least it announced and 
produced a great astronomer." Brahe's book De Nova Stella (1573), in 
which he first set down his observations and the conclusion that he 
drew "about the new star," caused immense interest and some horror 
at what were seen as sensational ideas. 

In De Nova Stella, Brahe described his first sight of the supernova: 

Last year in the month of November, on the nth day of that month, in the 
evening, after sunset, when according to my habit, I was contemplating the 
stars in a clear sky, I noticed that a new and unusual star, surpassing the 
other stars in brilliancy, was shining almost directly above my head; and 
since I had, almost from boyhood, known all the stars of the heavens perfectly 
{there is no great difficulty in attaining that knowledge), it was quite evident 
to me that there had never before been any star in that place in the sky, even 
the smallest, to say nothing of a star so conspicuously bright as this. 

This cool account is somewhat at odds with Brahe's further admission 
that, doubtful of the evidence of his eyes, he sought confirmation from 
his servants and some peasants driving by that they too could see the 
new star. They could. 

The unmoving star 

Brahe's most important measurements of the supernova of 
1572 were of its position. Although he did not have the advantage of the 
more accurate instruments which he later acquired for his observatory 
on the island of Hveen, he did have a large and well-made sextant-type 
instrument which he had just finished making. He was able to measure 
the distance of the supernova from the nine principal stars of Cassiopeia, 
making measurements as accurately as possible with the naked eye. He 
repeated the measurements at every opportunity, often several times 
throughout the night. In fact, the star was sufficiently near to the sky's 



j./£ tdptitCtfiiope* 

B pctltts Schedir. 

C Cmgulum 



G fuprma Ctthtirt 

H mtdid Chttcdr* 

I NotuiltlU. 


Ty cud's map of the 1572 supernova shows the brighter stars of 
Cassiopeia with the new star, marked " /," brilliantly outshining 
them. His Latin labels identify the stars by their positions in the 
mythological " Lady in the Chair" that the constellation is 
supposed to represent. Thus "A" is the head, "£" the knee and 
so on. 

north pole, the star Polaris, so that from Denmark the star never set 
and could be kept under observation the year round. Like Mastlin and 
Digges, Brahe found that its position was unchanged for all this time, 
from hour to hour, from day to day, from month to month to within the 
accuracy of his sextant. From other measurements of the positions of 
stars, we know that Brahe's measurements repeated to an accuracy of a 
few minutes of arc or less than a tenth of a degree (about the size of a 
nickel or a British penny held at a distance of 40 yards). Brahe's measure- 



ments firmly put the supernova unmovingly among the other fixed stars. 

What motion might Brahe have expected? It was natural for 16th 
century astronomers to compare the supernova with other transient 
phenomena, such as comets. Comets move, characteristically right 
across the celestial sky in a few months or even weeks. If the new star 
had been moving at such a rate, Brahe would have detected its motion 
in a matter of hours, unless, as some of his contemporaries implausibly 
argued, it was moving directly away from or towards Earth. 

He detected no motion in 18 months, eliminating any idea that the 
new star might have been associated with a planet, since the farthest then 
known, Saturn, would have moved with motion detectable by Brahe in a 
week. These general arguments, as well as the observation that the star 
twinkled just like other fixed stars, in contrast to the planets which shine 
without twinkling, were used by Brahe to show that the supernova was 
indeed a star in the eighth sphere, the firmament. But Brahe had 
specific arguments to show how distant the star was. His measurements 
with his sextant showed that the star had no parallax, or apparent 
motion caused by the motion of the Earth, and was certainly beyond the 
sphere of the Moon. Let us see how he was able to prove this. We will 
give the argument in terms of the Earth's rotation, although Brahe would 
have assumed the Earth to be stationary, and the firmament to be 

Brahe measured the position of the supernova from Heridsvaad when 
the star was almost overhead at the zenith in the evening sky (Fig. i). 
12 hours later, as morning approached, the Earth had rotated halfway 
round and, since the supernova was circumpolar and could be seen all 
night, Brahe was able to repeat his measurement. 

Brahe would be making the second measurement from a new position 
in space carried there by the rotation of the Earth. If the supernova were 
close to the Earth, as drawn in Fig. i, it would no longer be in the same 
direction. The distance of the supernova is measured in fact by the 
angle by which it shifts its apparent position, the so-called parallax, 
of the supernova. The smaller the parallax, the more distant the star. 



Parallax of 
I the supernova 

' \ Parallax of 


d£ ,«• 

Position of 
Heridsvaad at 
two times twelve 
hours apart 

Brake's measurement of the parallax of the supernova showed that 
it was less than the parallax of the Moon and hence the supernova 
was more distant. 



Brahe calculated that at the distance of the Moon the parallax of the 
supernova would be about a degree, whereas from his measurements its 
parallax could not be in excess of a few arc minutes, making the super- 
nova at least ten times as far away as the Moon. 

Five years later, Tycho was also able to show that the comet of 1 577 
had no parallax either, but traveled across the solar system, passing 
unhindered through any supposed crystal spheres that carried the 
planets, and proving that further dramatic changes took place above the 
Moon's orbit. 

Other astronomers than Brahe had suspected that the supernova of 
1572 had no parallax and must be among the fixed stars. Brahe's 
accurate measurements proved it, and that the celestial regions above 
the Moon were not as unchanging as Aristotle had thought. But Brahe 
was unable to bring himself to state this conclusion in his book De Nova 
Stella. On the contrary, he writes that all philosophers agree "that in the 
ethereal region of the celestial world no change of generation or of 
corruption occurs . . . but celestial bodies always remain the same, like 
unto themselves in every way." He argued that God had concealed the 
supernova from earthly eyes since the creation, choosing to reveal it 
when He wished. 

Brahe's cautious conservatism, which may have stood him in good 
stead in making careful observations, perhaps made him too unimagina- 
tive to perceive their consequences. He interpreted his observations of 
planetary motion in terms of epicycles and of his supernova in terms 
of suddenly revealed immutability. His pupil Kepler overthrew both 

Kepler's Supernova of 1604 

Tycho Brahe died in 1601, and his work was continued and 
developed by his former assistant, Johannes Kepler, a German. When the 
next supernova appeared in 1604, Kepler was working in Prague as 
court mathematician and tame astrologer for the erratic, and probably 
mad, Holy Roman Emperor Rudolf II. 



In September 1604, many eyes were turned towards the region of sky 
in which Mars and Jupiter were slowly drawing together — a sight which 
would attract the attention even of non-astronomers. The supernova 
appeared in the nearby constellation of Ophiuchus and, thanks to the 
conjunction, observers saw the supernova when it first appeared and 
when its brightness was still increasing. It is rare for a nova of any kind 
to be discovered before it is at its maximum light, because the increase 
from obscurity to full brilliance is so rapid. 

On October 9, two Italians saw the new star. One was an anonymous 
physician in Calabria, who reported what he had seen to the astronomer 
Clavius in Rome. The other was the astronomer I. Altobelli in Vienna. 
On October 10, a court official in Prague, J. Brunowsky, caught a glimpse 
of the new star between clouds and notified Kepler. 

Unfortunately, the weather in Prague was cloudy from then until 
October 17, but some observations were made from places where the 
skies were clear. When Kepler did see the star on October 17, it was very 
striking. He wrote that it competed with Jupiter in brilliance and that 



it was colored like a diamond. This proved to be near the date of maxi- 
mum brightness for the supernova as all observers agreed that there was 
no further increase in brightness after October 15. Kepler made arrange- 
ments for continued observations to be made, but in November the 
supernova was too close to the Sun to be seen. It reappeared from behind 
the Sun in January 1605, by which time it was already fading. It con- 
tinued to be visible until October 1605, and was carefully observed by 
Kepler and others until then. When the Ophiuchus region came out 
from behind the Sun in the spring once again, the supernova had become 
invisible to the naked eye. 

The European astronomers were not alone in observing the supernova. 
It appears in Chinese and Korean records as well. Apart from their 
interest in their own right, the Chinese observations are important 
because they can be compared with the European results and enable us 
to assess the accuracy of other Chinese records, such as those of the 1054 
supernova. The 1604 star occurs as one of the last entries in a list of 
guest stars in a document known as the She-ke. The entry is as follows : 

In the 32nd year of the same epoch, the gth moon, Yih Chow [October 10, 
1604] a star was seen in the degrees of the stellar division Wei. It resembled 
a round ball. Its color was reddish yellow. It was seen in the southwest until 
theioth moon[October 2/-November 26, 1604] when it was no longer visible. 
In the 12th moon, day Sin Yew [February 3, 1605] it again appeared in the 
southeast, in the stellar division Wei. The next year in the second moon 
[March 24- April 23], it gradually faded away. In the 8th moon, day Ting 
Maou [October jth, 1605] it disappeared. 

This agrees entirely with the European records and indicates the 
accuracy of the Chinese observations. Korean records also mention the 
supernova, describing it as having a greater magnitude than Jupiter, 
being a reddish-yellow color and scintillating. All European observers 
in October 1604 remark upon the red, orange or yellow color. 

Because of the philosophical controversies raging in 17th century 



Europe one of the main reasons for the great interest which the supernova 
aroused was the question of whether or not the star, like Brahe's super- 
nova 30 years before, was among the fixed stars. For this reason, great 
attention was given to measurement of its position in order to determine 
whether or not it had parallax. It had none large enough to be measured. 
Kepler produced a position for it but his figures are less satisfactory than 
those of David Fabricius at Osteel. From Fabricius's figures, modern 
astronomers have been able to compute the position to within one 
minute of arc. 

The supernova caused great interest and much speculation. Kepler 
said that it would bring good fortune to publishers at least, as it would 
bring a spate of pamphlets and books. He himself at once rushed into 
print with an eight page pamphlet in German describing the star and 
comparing it with Tycho's star of 1572. In 1606 he published his book 
on the subject, De Nova Stella, which was, he said, "a book full of 
astronomical, physical, metaphysical, meteorological and astrological 
discussions, glorious and unusual." As this description makes clear, the 
book was by no means as astronomical paper as we understand the term, 
but it does contain astronomy. The Emperor Rudolf IPs interest in 
astrology was such that Kepler was forced to include a great deal of 
interpretation and prediction. His main advice shows restraint: we 
should all consider what sins we have committed and pray for forgive- 

Speculations about the supernova 

The burning question to many, however, was this: was the 
star among the fixed stars and therefore another indication that there 
was change above the Moon? Kepler's answer was that, like Brahe's 
supernova, and for exactly the same reasons, the star was indeed above 
the Moon. Aristotle's model universe had failed in an important 
respect, and the rest of it was now under suspicion. Each of its attributes 
had to be subjected to scrutiny and tested against observation of the 



real world. Science became not a question of "saving the appearances," 
and making small modifications to an agreed-upon philosophy, but of 
attacking the basis of the philosophy itself. 

Kepler went on to use Tycho's observations of the positions of the 
planets to determine that the orbits of the planets were not perfect 
circles at all, but flattened, forming ellipses. Aristotle's concept of 
perfection in the regions above the Moon was false in every detail. 

Another controversial question concerned the origin of the star. One 
theory being discussed was that the conjunction of the planets had 
created such fire that they had ignited it. Kepler did not accept this idea. 
He believed that there is a material scattered through space which has 
an inherent ability to gather and then ignite itself. In favor of this idea, 
he argued that certain simple life forms appear spontaneously, in line 
with the common belief at the time that maggots appeared spontaneously 
in dead flesh. Shakespeare's Hamlet refers to this belief: "For if the sun 
breed maggots in a dead dog . . . ." He implies that Ophelia may con- 
ceive (though not necessarily spontaneously). Although Hamlet uses the 
idea ironically, it was a serious belief as Kepler's argument shows. 

Controversy over the supernova was not confined to Prague where 
Kepler was working. In Italy, at Padua, Galileo Galilei was professor of 
mathematics at the university and had won a reputation for his brilliant 
lectures and treatises on mathematics and mechanics. Galileo's interest 
in astronomy, however, did not come to the fore until after 1609 when 
he made his first telescope. Nevertheless, in 1604 he was acquainted 
with the problems of astronomy. Tycho Brahe himself had written to 
Galileo in 1600 and invited him to enter into a scientific correspondence. 
Galileo seems to have snubbed Brahe and he certainly never entered into 
extensive correspondence with him. Later, Galileo showed himself 
extremely hostile to Brahe's ideas, particularly his system of the Uni- 

When excitement arose over the 1604 supernova, Galileo, as a leading 
scientist, was asked to make a statement just as modern Nobel prize- 
winners are invited by the media to comment on scientific discoveries of 



which they often know little. Galileo was subjected to some criticism by 
his sponsors, the Padua city council, because he had not discovered the 
supernova, to which he somewhat peevishly replied that he had more 
important things to do than gaze out of the window, on the off-chance 
that he might see something interesting. Galileo was particularly pressed 
on the question of whether the star was superlunary or merely a 
meteorological phenomenon as the Aristotelians claimed. 

Galileo seems to have been reluctant to make any statement. One 
reason for his scientific stature was that he required evidence from 
first hand observations before he would offer interpretations. He 
expressed his attitude clearly in a parable about a man who understood 
well the technicalities of how to produce musical notes, but when he 
held a singing cricket in his hand had no idea how it produced its song. 
Galileo went on: 

The less people know and understand about such matters the more positively 
do they attempt to reason about them, and on the other hand, the number of 
things known and understood renders them more cautious in passing judge- 
ment about anything new. 

When he did give three lectures on the supernova to overflow crowds 
in the largest hall in Padua, as far as is known he made no definite 
statement on one side or the other. No texts for the lectures have 
survived, but from the indirect evidence available he seems merely to 
have stated the case that had been so far made by each side, and he 
declined to draw conclusions. Apparently he began a book on the sub- 
ject, but he never published it and only a small fraction of the manu- 
script remains. For the time being, he left the issue in doubt. 

Explosions in the minds of men 

The philosophical conclusions initiated by the supernovae 
of 1572 and 1604 and fully confirmed by Kepler's planetary theories, 
revolutionized the science of astronomy. But the new stars exploded in 



the minds of non-astronomers as well as in the sky. The first attempts 
to understand them were astrological. Brahe speculated that a period of 
peace to be followed by years of violence was indicated by the clear, 
white light of the star of 1572 which had been followed by a red, 
martial light. There were many similar predictions. Theodore Beza, a 
French Protestant theologian, wrote a Latin poem suggesting that the 
star was the Star of Bethlehem, implying that it heralded the Second 
Coming of Christ. Queen Elizabeth I of England sent for a leading 
astrologer, Thomas Allen, to ask the meaning of the star, to which "he 
gave his opinion very learnedly." But as it had been established early on 
by Mastlin and Brahe that the star was not a comet, astrologers were 
being confronted with a phenomenon for which there was no clear 
European precedent. Sir Thomas Browne considered that the whole 
business of predictions by astrology had been brought into disrepute : 

We need not be appalled by Blazing Stars, and a Comet is no more ground 
for Astrological presages than a flaming chimney. 

From many writers of the 17th century, we receive an impression of 
doubt and uncertainty. Astronomers themselves were disputing the 
implications of the lack of significant parallax of the supernovae; what 
was a layman to think ? Richard Corbet, Bishop of Oxford, in his letter 
to Master Ailesbury, written in 1618, expresses this uncertainty: 

tell us what to trust to; ere we wax 
All stiff and stupid with this Paralax. 
Say shall the old Philosophy be true, 
Or doth He ride above the Moon think you? 

If Corbet sounds unenthusiastic for the new problems of parallax, 
Henry More, a Cambridge scholar, writing later in 1642 is sceptical and 
uncomplimentary to those who still held to the old philosophy : 



That famous star naiVd down in Cassiope 
Horn was it hammer 'd in your solid sky? 
What pinsers pulVd it out again that we 
No longer see it, whither did it fly? 

Probably the most famous of the expressions of discomfort at the new 
ideas is from the 17th century poet, John Donne. His poem, An Anatomie 
of the World shows clearly how the discovery of supernovae could help 
to bring the structure of his whole world tumbling about a man's ears. 

And new philosophy calls all in doubt . . . 
And freely men confesse that this worWs spent, 
When in the planets and the firmament 
They seeke so many new. 

Here, as in many of his comments on astronomy, Donne is accusing in 
his tone, implying that astronomers are forcing new worlds on our 
attention with all the new ideas that they imply, rather than merely 
observing them and reporting on what they observe. Donne points out 
the full logical conclusion which must be drawn from abandoning the 
old concept of the Universe. If the order in the heavens was gone, what 
about order in the state, the church, the family ? 

Tis all in peeces all cohaerence gone; 

All just supply and all Relation; 

Prince, Subject, Father, Sonne, are things forgot? 

Reluctance to abandon all the old certainties is not surprising. At the 
same time, the "new stars" appealed to an age of explorers. There was 
some fitness that when the map of the Earth was rapidly being drawn 
larger and larger, there should be greater knowledge of the sky. Unlike 
Brahe, Donne implied that seeing a supernova is discovering a new star, 



rather than seeing an event which by its nature could not have been seen 
before. This attitude is part of his general hostility at this point to the 
"new philosophy." Likening America to the supernovae he thought that 
both would have the same effect on men's minds. 

We have added to the world Virginia and sent 
Two new starres lately to the firmament. 

He addresses astronomers and explorers in one breath, 

You, which beyond that heaven which was most high 
Have found new spheares and of new lands can write . . . 

To Donne personally the new knowledge was cause for regret. There is 
no mistaking the poignancy in speaking of "heaven which was most 
high." Yet his frequent references to "new starres" show the impact 
that astronomy at that time would have on a man who thought about 
its implications. 

For most people, the new stars co-existed peacefully beside their 
religion. To the compartmentalized or unquestioning mind, there were 
no uncomfortable religious or philosophical doubts. A supernova was 
merely a more than usually beautiful star. The poet Edmund Spenser 
tells a new bride to shine like a supernova. 

Bee thou a new starre that to us pertends 
Ends of much wonder. 

John Dryden used Tycho himself in a poem as a means of praising 
Lord Hastings. 

Liv'd Tycho now, struck with this Ray {which shone 
More bright P the Moon than others' Beam at Noon) 
He'd take his Astrolabe and seek out here 
What new Star 'twas did gild our Hemisphere. 



There are no philosophical questionings, simply an acceptance of the 
supernova as a beautiful phenomenon which was part of the educated 
man's horizon at the time. 

Tycho himself had compared the observation of the 1572 supernova 
with the stopping of the sun by Joshua or the Crucifixion, in its momen- 
tous effect on the mind and imagination. One feels that he would have 
been pleased by the wealth of writing in which his supernova was 

The questions which the two stars of 1572 and 1604 had raised, 
however, were to be answered finally and decisively only after the 
development of a new technology: the telescope. 

Answers from the telescope 

Historians are still arguing about who invented the 
telescope. Some say it was a Dutch optician, Hans Lipperhey, around 
1608, while others maintain that the instrument had already been known 
for a decade or more before Lipperhey was granted a license to manu- 
facture telescopes. Similarly, there is considerable evidence that 
Galileo was not the only person to turn a telescope on the stars in 1609, 
but what is certain is that it was Galileo who had the power of intellect 
to understand the implications of what he was seeing, and to relate that 
to the recent supernovae. 

After experiments in producing several prototypes, Galileo designed 
a satisfactory telescope which gave him a clear view of the sky. Naturally, 
he pointed his telescope to well-known objects so that he could reveal 
new aspects of them. He found craters on the Moon, satellites orbiting 
Jupiter, rings around Saturn, spots on the Sun, new stars in the Pleiades 
star cluster . . . the list goes on and on. His discoveries revolutionized 

He recorded his discoveries in The Starry Messenger (16 10). Between 
1623 and 1 63 1 Galileo summarized his cosmology in his Dialogue con- 
cerning Two World Systems. The book is written as if it were a three way 



conversation between an Aristotelian and a Copernican philosopher who 
try to convince an uncommitted disputant of the truth of their cause. 
Galileo puts his own view on the Universe through the arguments of the 
Copernican, Salviati, but he had been forced by the Pope to promise 
to give equal weight to the official Catholic position as well. He did this 
through the mouth of a character unsubtly named Simplicius and it is 
fairly obvious that Galileo's sympathies lay with the other debater, 
Salviati. One of the characters asks how astronomers could tell whether 
the new stars were "very remote." To this Salviati replies, 

Either of two sorts of observations, both very simple, easy and correct, 
mould be enough to assure them of the star being located in the firmament, 
or at least a long way beyond the Moon. One of these is the equality — or 
very slight disparity — of its distances from the pole when at its lowest point 
on the meridian and at its highest [the measurements made by Brake, and 
described on p. 36]. The other is that it remained always at the same distance 
from certain surrounding fixed stars; especially Kappa Cassiopeiae, from 
which it [the i$j2 supernova] was less than one and one half degree distant. 
From these two things it may unquestionably be deduced that parallax was 
either entirely lacking or was so small that the most cursory calculation 
proves the star to have been a great distance from the Earth. 

The weight of evidence was now such that, as Galileo says, if Aristotle 
had been alive he would have changed his mind about the immutability 
of the heavens. Spots had been seen on the face of the Sun. Comets have 
been observed which have been 

generated and dissolved in parts higher than the Lunar Orb, besides the two 
new stars, Anno 1572 and Anno 1604, without contradiction much higher 
than all the planets. 

Nevertheless, Galileo had to pay the high price of a summons by the 
Inquisition and a trial in Rome for stating these views, which were 



unacceptable because they obliterated the distinction between the 
corruptible and incorruptible, placing the Earth among the heavens and 
bringing the heavens down to Earth. 

Galileo suffered because his discoveries with his telescope made it 
impossible to maintain a belief in the unchanging heavens. He found 
himself obliged to state the conclusions which had in fact been inevitable 
since 1572. The supernovae had changed European cosmology. 

What they mean to 20th century cosmology is equally momentous 
but fortunately less likely to arouse animosity or fear. 


IV Supernovae in other galaxies: 
supernovae sought 


.here has been, on average, a bright supernova visible in 
our Galaxy every 200 years. Unfortunately, not one has been seen since 
Kepler's in 1604. Astronomers have had to move gradually towards a 
better understanding of what supernovae are from observations of 
supernovae in other galaxies. When extragalactic supernovae were first 
seen they also caused incredulity and controversy, and when the modern 
founder of supernova research, Fritz Zwicky, began his systematic 
search for supernovae in other galaxies he was accompanied, he says, 
"by the hilarious laughter of most professional astronomers and my 
colleagues at Caltech." But before the first supernova was found by 
deliberate search, the understanding of supernovae continued to grow, 
helped by various chance discoveries. 

A highly remarkable change 

On the evening of August 20, 1885, E. Hartwig of the 
Dorpat Observatory in Russia was discussing the Laplace theory of the 
origin of the planets with friends. In broad outline this theory, put 
forward by Pierre-Simon Laplace in 1796, is similar to that widely 
accepted today. 

The starting point of the theory is a huge slowly-rotating gas cloud. 



As the cloud contracts its rotation speeds up until it becomes fast enough 
to throw off rings of material which then condense into planets. The 
central part of the cloud eventually becomes the Sun. 

Around 1885, when Hartwig's philosophical discussion took place, 
astronomers were trying to link Laplace's theory with the observational 
fact that many nebulae were being discovered to have a spiral shape, 
though none had been resolved into stars. It was not surprising that 
many people thought them to be planetary systems in formation within 
our own Galaxy. 

As Agnes Clerke, the British astronomer-writer, wrote with rash 
certainty in 1890: "No competent thinker, with the whole of the 
available evidence before him, can now, it is safe to say, maintain any 
single nebula to be a star system of coordinate rank within the Milky 

Hartwig's friends must have been eager to see one of these mysterious 
nebulae through a telescope, because he took them out to look at what 
was then called the Great Nebula in Andromeda, Messier 31. Through 
Dorpat Observatory's 9-inch refractor very little of interest would then 
have been visible, since the Moon was approaching Full with its bright 
light swamping the delicate structure in the nebula. Hartwig had sur- 
veyed M31 on three occasions during the previous New Moon period, 
and must therefore have been astounded to see at the center of M31 a 
new star shining brightly where no star had shone before. Clearly, 
thought Hartwig, this was a central sun appearing as predicted by 
Laplace's theory. 

The star was seen in Hungary by Baroness Podmaniczky on August 22 
or 23. She failed to realize its importance. At Heidelberg it was seen by 
Max Wolf on the 25th and 27th while testing a telescope, but he thought 
the star was an effect of moonlight. In Rouen, Ludovic Gully at a public 
night at the newly opened observatory on August 17 had seen the star 
in the new coude telescope, but the telescope had been giving trouble 
in its tests and Gully thought that the appearance of the star was caused 
by a defect. 



Only Hartwig saw the significance of the new star, presumably because 
he had been studying the region just the week previously. He could not 
convince the observatory's director of the new star's reality, however, 
and was not allowed to telegraph the discovery to the central clearing 
house for astronomical information in Kiel. Hartwig wrote to Kiel 
anyway but the letters went astray because of the petty theft of the 
stamps from the envelopes. He was not allowed to announce the "highly 
remarkable change in the Great Andromeda Nebula" until he and the 
director had confirmed the existence of the new star in the moonless 
sky on August 27. By that time the star had already noticeably faded. 
Hartwig was able to follow it almost daily as it declined. 180 days after 
its maximum, which had probably been on August 17, it was beyond 
the largest telescope's light-gathering power. 

The nova was used in renewed attempts to decide whether or not M31 
was a galaxy or a nebula. It had become clear that astrophysical arguments 
were too inconclusive because theories of the behavior of stars and 
interstellar gas were not well enough developed. For example, in 1899 
J. Scheiner found that M31 had the spectrum of a collection of sunlike 
stars rather than the spectrum to be found in hot gaseous bodies. 
Scheiner correctly inferred that M31 was an aggregate of distant stars. 
But when V. M. Slipher examined the gas in the Pleiades star cluster, 
he found that it too had a starlike spectrum, and thought that the 
Andromeda Nebula might be the same. We now know that the Pleiades 
gas is dust-laden and simply reflects the light of the bright Pleiades stars. 

Clearly only astronomical fact could determine the truth, and the 
prime fact required was the distance of M31. If M31 was no more distant 
than the stars of the Milky Way, it could not, in the picturesque phrase 
of the time, be an "island universe" of stars. 

In the determination of the distance of M31 the nova observed by 
Hartwig, and now named S Andromedae, played a highly misleading 

5 2 

A misleading comparison 

In 191 1 F. W. Very compared S Andromedae with a nova 
which occurred in the Milky Way in Perseus in 1901. Nova Persei 1901 
brightened in 28 hours from invisibility to naked eye brightness. Three 
days later, when it was at its brightest, it was among the half dozen 
brightest stars. It then faded over a few months back to invisibility. 

Within six months Max Wolf noticed on photographs of the region 
of the nova that it had become surrounded by a small nebula. Because 
this was so soon after the nova outburst, astronomers realized that this 
nebula could not be gas ejected from the nova. Instead, this must be a 
reflection nebula. As the burst of light from the star spread at the speed 
of light into the surrounding space it illuminated a hitherto invisible 
dark nebula, reflecting light from its successive layers. As the nebula 
expanded, its brightness diminished: after two years it was a dim patch 
larger than the Moon and it gradually faded away. 

Comparing the size of the nebula at various times with the speed of 
light at which it had been formed, astronomers were able to determine 
the distance to Nova Persei 1901 as some 500 light years. 

At its maximum, therefore, Nova Persei was about 250 times brighter 
than S Andromedae at its maximum. Because the dimming of the light 
of distant objects depends on their distance squared, it followed that if 
Nova Persei had been moved about 16 times farther away it would have 
appeared to be 250 times fainter, the same as S Andromedae. Therefore, 
if S Andromedae and Nova Persei were alike, the Andromeda Nebula 
could be no farther than 8000 light years from Earth, not a very large 
distance and still well within our own Milky Way. From this, F. W. Very 
argued that M31 was not another galaxy like our own. 

The discovery of further novae in spiral nebulae rekindled interest 
in this argument in 1917. G. W. Ritchey had photographed a nova in 
another spiral, catalog number ngc 6946, and this inspired him to 
re-examine all the photographs of spiral nebulae, including M31, taken 
by the Mount Wilson 60-inch telescope since 1908. Among the six 



novae which he found were two in M31 that had passed unnoticed 
because they were much fainter than S Andromedae. Repeated photog- 
raphy of M31 by Ritchey, Harlow Shapley, John Duncan and R. F. 
Sanford quickly threw up eight further examples in two years, but none 
nearly as bright as S Andromedae. 

It began to be clear that S Andromedae was not a typical nova in M31, 
and that, in making comparisons between novae in our Milky Way and 
in the Andromeda spiral nebula, S Andromedae should be ignored. 
Realizing instantly that the "occurrence of these new stars in spirals 
must be regarded as having a very definite bearing on the 'island uni- 
verse' theory of the constitution of the spiral nebulae," H. D. Curtis 
went on to point out that the difference of brightness between novae 
occurring in our Milky Way and those in spiral nebulae, particularly 
M31, if S Andromedae were ignored, could be accounted for if the novae 
in M31 were 100 times the distance of Milky Way novae — "that is," 
writes Curtis, "the spirals containing the novae are far outside our 
stellar system." Curtis thus found himself a protagonist of the theory 
that spiral nebulae were galaxies like our own, "island universes" 
independent and separated from our Milky Way. 

The Great Debate begins 

In the same issue of the Publications of the Astronomical 
Society of the Pacific (October 1917) Harlow Shapley developed the 
same numerical argument as Curtis, but highlighted the problem of 
S Andromedae. He pointed out that it must have had a luminosity 
100 million times that of the Sun, and that "this remarkable result must 
inevitably follow if spiral nebulae are considered external galactic 
systems comparable with our own in size and constituency." 

During this period Shapley was studying the stars in dense spherical 
clusters (globular clusters), making a stellar census, and by 1919 had come 
to the conclusion that the study of globular clusters had yielded sufficient 
knowledge of the luminosity of more than a million stars to show that 
not one was anywhere near the enormous brightness of S Andromedae. 



Hence, he argued, stellar luminosities of the order ioo million times that 
of the Sun seemed out of the question, and accordingly the close 
comparability of spirals containing such novae to our galaxy appeared 
inadmissable. Shapley found himself on the other side of the debate 
from Curtis, opposing the "island universe" hypothesis of the spiral 

A formal Great Debate on this subject was held on April 26, 1920 at 
the National Academy of Sciences in Washington, D.C. Curtis was 
forced to grasp the nettle of S Andromedae and to conclude that it 
seemed certain that the range of brightness of the novae in the spirals, 
and probably also in our Galaxy, may be very large, as is evidenced by 
a comparison of S Andromedae with the faint novae found in M31. A 
division into two classes was not impossible. With this statement Curtis 
ventured for the first time the concept that, besides ordinary novae, 
there exists a class of much brighter novae. 

The Great Debate itself, of which the argument about the novae was 
a part, was inconclusive. Neither astronomer convinced the other, nor 
could their contemporaries decide what was the true status of the spirals. 
Not until 1923 did Edwin Hubble identify in M31 examples of a type of 
variable star called cepheids, using the greater power of the 100-inch 
Mount Wilson telescope to measure the varying brightness of these 
faint stars. Comparison of these stars with their Milky Way counterparts 
proved that, according to Hubble's figures, M31 was some 1 million light 
years away and was indeed a so-called "island universe" external to our 
own Galaxy. 

Curtis' view of the status of the spirals was shown to be nearest the 
truth, and the Great Debate has been finally resolved in his favor. 

Supernovae revealed 

As the evidence began to be assimilated, the startling 
brightness of S Andromedae became more apparent. Since M31 is a 
galaxy like our own it contains stars by the billion. S Andromedae at 
maximum brightness was equal to one sixth of the light from the entire 



galaxy! Edwin Hubble himself recognized in 1929 this consequence of 
his observations of the cepheids in M31 : S Andromedae belongs to "that 
mysterious class of exceptional novae which attain luminosities that are 
respectable fractions of the total luminosity of the system in which they 

Up to 1933 haphazard photography of galaxies had thrown up a total 
of 19 examples of new stars having the property that they nearly equaled 
the brightness of the galaxies in which they were found. Walter Baade 
and Fritz Zwicky christened these stars supernovae and outlined their 
fundamental properties. To them belongs the credit for being the first 
to understand the key part that supernovae were to play in modern 
astronomy. In its entirety the summary of their paper, translated into 
non-technical language, reads as follows: 

Supernovae flare up in every galaxy once in several centuries. The lifetime 
of a supernova is about 20 days and its brightness at maximum may be as 
high as 100 million times that of the Sun. Calculations indicate that the 
total radiation, visible and invisible, is about 10 million times what can be 
seen. The supernova therefore emits during its life a total energy equal to 
the amount that the Sun would radiate in a million years. If supernovae 
are initially quite ordinary stars of mass up to about 10 times that of the 
Sun, the amount of energy they release is comparable to the energy that 
would be made if their mass was turned directly to energy. Therefore in the 
supernova process mass in bulk is annihilated. 

In addition, the hypothesis that cosmic rays are produced by supernovae 
suggests itself. Assuming that in every galaxy one supernova occurs every 
thousand years the intensity of cosmic rays expected to be observed on Earth 
is equal to the level actually observed. With all reserve we advance the view 
that supernovae represent the transitions from ordinary stars into neutron 
stars which in their final stages consist of extremely closely-packed neutrons. 

In a single 15-line summary, fully aware of the bizarre nature of 
what they were saying, Baade and Zwicky mapped out the achievements 



of the next 40 years' work on supernovae, as the details of their outline 
were filled in. 

Zwicky himself recalled that when in 1934 he bought a camera just 
to prove that he and Baade were right by photographing the rich cluster 
of galaxies towards the constellation Virgo from the top of a building of 
the California Institute of Technology he was scorned by his colleagues. 
Expecting to find two or three supernovae in two years he in fact found 
none. Though he was almost inclined to give up the systematic search 
for supernovae because of this setback, Zwicky persuaded George 
Ellery Hale, Director of the Mount Wilson Observatory, to divert some 
money from the grant from the Rockefeller Foundation to build the . 
200-inch telescope in order to construct a new type of camera which 
would better enable him to find supernovae in other galaxies. 

Bernhard Schmidt, an Estonian optical designer, had recently sug- 
gested a way of giving an astronomical camera-telescope a much larger 
field of view, using a corrector lens. Instead of the usual half- degree or so, 
the new Schmidt camera photographed a circle of sky eight degrees 
across and had an aperture of 18 inches — an ideal instrument for looking 
at many galaxies at once. Zwicky got his telescope. 

In the period from September 1936 to December 1939, Zwicky and 
his co-worker J. J. Johnson took 1625 photographs of 175 regions of the 
sky chosen to contain nearby galaxies. Taking into account the number 
of galaxies in the field of the telescope the program amounted to 5150 
years continuous observation of an average galaxy. Twelve supernovae 
were discovered in those three years, giving a rate of one discovered 
supernova every 430 years per average galaxy. The second which Zwicky 
discovered, at maximum brightness on August 22, 1937, was more than 
100 times brighter than the total light of the galaxy in which it appeared, 
an irregular spiral galaxy called 1 c 4182, and was the brightest supernova 
seen so far in this century. 

During World War II the search essentially stopped, but Zwicky 
persuaded Hale to ask the Rockefeller Foundation for nearly $500,000 
to construct a larger 48-inch Schmidt telescope, put into operation in 


A supernova flares in a distant 
galaxy. The top picture shows the 
galaxy NGC 5253 as it usually 
appears on photos taken with the 
48-inch Palomar Schmidt. 

January 4, 1959 

May 6, 1972 June io, 1972 

1949. This telescope's first job was to photograph the entire sky in the 
Palomar Sky Survey, which has been a cornerstone of astronomy for 
many years; but the Palomar Supernova Search was resumed in 1958 
and is still under way, directed by Wallace Sargent and Leonard Searle. 
The man who carries out the photography and who makes the discoveries 
is Charles T. Kowal, who has compared himself to a ship's engineer, 
making the ship run while the captain walks the bridge. Three moonless 
nights per month are scheduled for the search, which is of 38 fields 
containing a total of 3003 galaxies in clusters and groups. The photo- 
graphs taken at night are compared the following afternoon, if possible, 
with standard photographs of the same fields, so that newly discovered 
supernovae can be rapidly followed up. Between 1958 and 1971, the 
search produced 63 supernovae, and the total discovered at Palomar 



The lower four show the 1972 supernova in this galaxy. The brightness of the 
supernova declined from its maximum (far left), when it was as bright as all 
the other stars in NGC 5253 combined, to near-invisibility after almost a year 
(far right). The fifth supernova to be discovered in an external galaxy in that 
year, igj2e was of Type I. It lies well from the center of its galaxy, in regions 
too faint to be recorded on these photographs. Hale Observatory photographs by 
Charles T. Kowal. 

January 30, 1973 April 24, 1973 

since the inception of Zwicky's search up to 1973 was 270 supernovae, 

at a cost, noted Zwicky, of $550 each. Other supernova searches are 

carried out from Asiago (Italy), Zimmerwald (Switzerland) and Konkoly 


The supernova birth rate 

It is difficult to determine the true rate at which supernovae 
occur in the galaxies searched. The statistics show that fewer supernovae 
are found in the fainter galaxies than in the brighter, presumably because 
the fainter galaxies are, on average, farther away, so supernovae in them 
will be correspondingly less noticeable. Supernovae are also missed 
when they merge with the central regions of the fainter galaxies. 



Sargent, Searle and Kowal have concluded that in a typical galaxy the 
average supernova frequency may be as high as one every 20 to 30 years, 
considerably more than Zwicky's first estimate. Up to his death in 
February 1974, Zwicky himself vigorously disputed such a high 

This high rate of occurrence of supernovae, even in our own Galaxy, 
is to some degree confirmed by studies of the number of pulsars. Radio 
astronomer Andrew Lyne has offered evidence based on a Jodrell Bank 
pulsar survey that the number of pulsars now visible suggests that they 
are created in our Galaxy at the rate of one every 10 to 20 years. If, as 
astronomers believe, pulsars are created in supernova explosions, the 
pulsar birth rate and the supernova birth rate should be the same. 

Where have all the supernovae gone ? 

The Sun lies almost exactly in the plane of the Galaxy — 
along the Galaxy's equator in effect — which is where supernovae 
mostly occur. If they occur as often as one every 20 years in our Galaxy 
why don't we see more? Why is it that astronomers have waited 300 
years without seeing a galactic supernova ? The culprit is the dust and 
other interstellar material which also lies along the galactic plane, 
obscuring much of what goes on even in our own neighborhood. Look 
out at the Milky Way on a dark moonless night. In Cygnus and Taurus 
the Milky Way appears cleft in two. In the Southern Cross a dark hole 
is silhouetted against the Milky Way. All these black patches are 
relatively near dark clouds of dust hiding the light of the stars behind. 
Calculations show that only 40 per cent of all supernovae that occur 
in the Galaxy can be visible to the naked eye, even at maximum. 
Apparently, over the last 1000 years, only ten per cent have reached a 
sufficient brightness for a long enough time, far enough from the Sun 
for their light not to be swamped by the dawn or evening sky, to make 
discovery possible. Modern astronomers cannot accept a 90 per cent 
failure rate, and have made special arrangements to catch the next 
galactic supernova. 


The Milky Way from the constellation of Scutum (top), to 
Scorpio at bottom. The horizon is blurred as the telescope taking 
this picture followed the stars. Silhouetted against the massed 
faint stars of the Milky Way are lanes of dark clouds which blot 
out the light from stars behind. Galactic supernovae preferentially 
occur along the center line of the Milky Way and are heavily 
obscured by the dark clouds. 

Looking for the next galactic supernova 

The last time anybody was able to stare up at the stars and 
see a supernova was in 1604. Despite the statistics, there has not been 
one visible to the unaided eye for almost 400 years. Obviously, the next 



time the Earth is illuminated by the glare of such a star, the attention of 
all astronomers will be focused on it. But whenever the next galactic 
supernova appears — and it could be tomorrow — the first problem will 
be to recognize it and distinguish it from ordinary novae. 

Ordinary novae in our Galaxy are seen to flare up at the rate of one 
or two per year. They are often picked out by watchful amateur astrono- 
mers who make a point of searching for them. The sky is so large, and 
nova searching so frequently fruitless, that professionals can find little 
time for this sort of thing. Amateurs, however, have the time and 
dedication to learn the appearance of the sky well beyond the limit of 
naked-eye vision. The game can be rewarding because, as with comets, 
new stars are sometimes named after their discoverer, who also has the 
satisfaction of knowing that the world's major telescopes will be turned 
towards his star. 

As a result of the keenness of the amateur nova patrol, many novae 
are now being spotted when they are too faint to be seen with the naked 
eye. This greatly increases the chance that even if a distant, heavily 
obscured supernova were to appear in the present day, it would still be 

Because the occurrence of such things as supernovae cannot be 
predicted and because they brighten up so quickly, astronomers have 
now set up an early warning system, run by the International Astro- 
nomical Union. Anyone — professional or amateur — discovering an 
important new object such as a nova or comet notifies the Central 
Bureau for Astronomical Telegrams in Cambridge, Mass., usually 
through the facilities of the nearest observatory. The Bureau then 
cables it subscribers, which include all principal observatories and 
amateur groups, using a brief and economical code, prefaced by the 
alerting codenames astrogram echo (for bright novae) or astro- 
gram France (for fainter ones). 

When an observatory receives one of the innocuous-looking strings 
of five-digit numbers, there is always a flurry of activity as astronomers 
decipher the groups and check charts of the right area of sky in their 



libraries. They then telephone details to colleagues at their observatory's 
telescopes. It is reasonably safe to predict that every suitable telescope 
will be pointing at it within 24 hours of the discovery of the next galactic 

If the next supernova occurs in the direction away from the center of 
the Galaxy, then it will almost certainly be bright enough to be visible 
to the naked eye. Most supernovae, however, are expected to lie towards 

After an initial brief telegram, the world's astronomers are given 
details of important discoveries by an I.A.U. Circular such as this 
one, airmailed rapidly to thousands of professional and amateur 
subscribers. This Circular, sent out May 18, igj2, announces 
KowaVs discovery of the supernova shown on pp. 58-^g. 


Circular No. 2405 





Mr. C. T. Kowal, Department of Astrophysics, California Insti- 
tute of Technology, telegraphs that he discovered on May 13 a 
supernova of magnitude 8.5 in NGC 5253 (a » 13 n 37 m l, 6 = -31°24\ 
equinox 1950.0). The object, located 56" west and 85" south of the 
nucleus, was confirmed on May 15. This seems to be the fourth 
brightest extragalactic supernova ever recorded; the second bright- 
est (= Z Cen), observed in 1895, was also in NGC 5253. 


Recent observations show that this object has been brightening 
again. C. E. Scovil, Stamford, Connecticut, gives the following 
magnitude estimates: May 5.17, 11.2; 6.24, 11.3; 7.11, 10.5; 11.28, 
9.6; 12.14, 9.6; 13.28, 9.7; 14.18, 9.6. P. Moore, Selsey, Sussex, 
England, gives: Apr. 29, 12.1; May 7, 10.7; May 10, 10.4. 


In calculating the following ephemeris (cf. IAUC 2330) the AT 
correction (as indicated by several semi -accurate observations by 
Dr. E. Roemer during July-December 1971) has been applied. 

1972/73 ET 





m 2 

June 12 

2 h 40™98 






2 59.65 

+24 16.9 

July 2 

3 18.70 

+25 43.7 





3 38.11 

+27 04.3 


3 57.77 

+28 18.0 




Aug. 1 

4 17.60 

+29 24.2 


4 37.46 

+30 22.2 





4 57.16 

+31 11.8 


5 16.50 

+31 53.1 




Sept. 10 

5 35.23 

+32 26.3 


5 53.01 

+32 52.1 





6 09.54 

+33 11.5 

Oct. 10 

6 24.39 

+33 25.6 





6 37.14 

+33 36.0 


6 47.33 

+33 43.6 




NOV. 9 

6 54.48 

+33 49.2 


6 58.21 

+33 52.6 





the galactic center, just because most of the Galaxy as seen from the 
solar system lies in that direction. If the supernova is relatively close to 
us, say at a distance of a few thousand light years, it will be possible to 
see the supernova grow in size as it expands into space. Only 40 days 
after it explodes a supernova at a distance of 3000 light years will appear 
non-stellar in a telescope. Possibly, if some theories of supernovae are 
correct, it will be large enough to be seen as a perceptible disk right at 
the time of maximum brightness even when observed by amateur 
astronomers with moderate-sized telescopes. To the naked eye the 
supernova may appear not to twinkle to the same degree as other bright 
stars, but might glow balefully. 

The burst of light from the supernova as it explodes will spread into 
the surrounding space, at the speed of light, and will illuminate any 
interstellar material that happens to lie near the supernova as was the 
case with Nova Persei 1901. Light from the supernova explosion will be 
reflected towards the Earth by previously unseen dark clouds of inter- 
stellar material and a ripple of light will be seen spreading out through 
interstellar material, away from the supernova. These light echoes may 
be visible for hundreds of years, expanding into space until the light is 
too diffuse to be seen. If they are visible to the naked eye, the light 
echoes will probably be detectable to a distance of one degree from the 
supernova, and the light echo will appear as a ring about four times the 

size of the Moon. 

If such a thing is seen surrounding the next galactic supernova, it will 
provide an easy way to determine its distance. One year after maximum, 
the radius of the light echo will be exactly one light year. Astronomers 
will be able to measure its angular distance at this time from the super- 
nova and determine, by trigonometry, the precise distance of the 

Although supernovae emit large numbers of X-rays and gamma rays, 
this high energy radiation will be prevented from reaching Earth by 
the dense expanding shell ejected by the explosion. There is possibly a 
brief moment at the beginning of a supernova explosion when there is 



some hope of detecting a blast of X-radiation, the so-called prompt 
emission, if there is an X-ray satellite orbiting the Earth at the time. A 
few years after a supernova explosion, the shell becomes thin enough to 
be transparent to X-rays again, and they could probably be detected 
even then if they are strong enough. 

As the shell ejected by a supernova disperses into space, it may cool 
to a point where dust grains can form. These will be warm when they 
first form and will emit infrared radiation. Therefore, at about the time 
when X-radiation becomes visible, infrared radiation will also be 
perceptible, while the supernova optically fades away. 

These ideas about the next galactic supernova are all based on a 
discussion presented by Sidney Van den Bergh in a paper intended to 
help astronomers to plan observations, should the event occur in their 
working lifetime. Van den Bergh remarks at the end of his article: 
"Those who might tend to become discouraged while they wait for this 
momentous occasion might be slightly consoled by the thought that the 
light of about 500 galactic supernovae that have already occurred is 
currently on its way to us!" 


V The Crab and its mysteries: 
a supernova remnant 

V Thile 

'hile waiting for the next bright supernova to 
study, astronomers have been studying the remnants of past supernovae, 
for their own interest and for the light they throw on the supernova 
phenomenon. Probably more effort has been put into understanding one 
particular remnant, the Crab Nebula, than any other astronomical 
object, save the Sun. Solving the mysteries of the Crab Nebula has 
progressed with the development of new astronomical instruments and 
techniques, starting with the invention of the telescope itself. Gradually, 
astronomers have come to understand how significant the tantalizing 
and infuriating Crab could be in understanding supernovae. No matter 
how many of its mysteries have been solved, the Crab seems always to 
hint at another problem. 

Like all the first-discovered nebulae, it was found by chance, not 
through a systematic search, because the invention of the telescope had 
posed astronomers a question which most, like Galileo, declined to 
tackle. How could they systematically survey the whole sky with the 
telescope and still have time for other studies ? Suppose we estimate the 
field of view of his telescope — the area the astronomer could see at any 
one time — at one degree (twice the diameter of the Moon). This figure 
is generous for telescopes of the 17th and 18th centuries. Over the whole 
sky there are 42,000 such areas of which perhaps a quarter are per- 



petually below the horizon from European latitudes. Allowing just four 
minutes for the inspection of such an area as the telescope, carried by 
the Earth's rotation, scans across it, we find that some 2000 hours are 
needed to sweep the telescope over the whole sky. Astronomical tele- 
scopes view the sky at night for about half the time (the other half is 
spent setting up the telescope, making notes, and so on) ; it is cloudy in 
Europe for half to two thirds of all nights; and half the time the Moon 
is too bright and floods the sky with light, washing out the fainter stars. 
Consequently, to be able to observe for 2000 hours in the best conditions 
requires about seven years, assuming that the astronomer is dedicated 
and is prepared to work throughout the year during all hours of darkness. 
Realistically, such an all-sky survey takes a substantial fraction of a 
working lifetime. Yet as telescopes have improved, astronomers have 
repeated examinations of the sky even to the present day. 

One such early astronomer was John Bevis (1693-1771), a Welsh 
doctor, who compiled the results of his observations of star positions into 
an atlas, the Uranographia Britannica. The costs of preparing the plates 
for the Atlas were so high that the printer went bankrupt and his creditors 
seized his assets including the engravings of the star charts. A few proofs 
of the Uranographia Britannica had been struck, however, and on these 
are plotted 16 nebulae (Latin for clouds), the term astronomers now 
use to refer to the unstarlike patches of light which they were beginning 
to discover. 

One set of proofs was shown to the French historian Lalande on a 
visit to London. Lalande, in a text on astronomy, records that the great 
French astronomer Charles Messier had another set. In 1758, Messier, 
like the whole of the astronomical community, was eagerly awaiting the 
appearance of Halley's Comet. Many years before, Edmond Halley had 
realized that this bright comet, then last seen in 1683, returns to the 
vicinity of the Earth and Sun every 75 years or so. Messier was known 
as the "ferret of comets" for his assiduous searches for and discoveries 
of new comets. He actually found a comet near the predicted place, but 
in fact it turned out not to be Halley's, which arrived later. Messier's 



new comet passed into the constellation Taurus, and in following it he 
chanced upon Bevis' nebula. In his own words: "The comet of 1758 
being between the horns of the Bull, I discovered on August 28 below 
the southern horn and a short distance from the star Zeta of that 
constellation, a whitish light, elongated in the form of a candle flame 
containing not one star." He compared the nebula with the comet and 
said that the nebula was more "vivid" and more elongated than the 
comet which seemed "almost round." 

This discovery was the first of many nebulae found by Messier in the 
course of sweeping the sky for comets. In 177 1 he published a catalog of 
all known nebulae with Bevis' nebula in Taurus in first place. After 
Bevis wrote to Messier pointing out that he had discovered the nebula 
first, Messier gave him the credit for it. Messier's catalog of nebulae is 
known by the initial of its compiler, and therefore Bevis' nebula is 
called mi. 

Although it is first in the catalog, mi is by no means the most 
prominent nebula in the sky. Others, such as the Orion Nebula (M42) 
are much more spectacular and can easily be seen with the naked eye. 
While Mi can be seen as a misty patch in fairly small telescopes under 
good conditions, it needs a moderately large instrument to show it well. 

Stars or gas ? 

As telescopes became better and better, astronomers began 
to find that many of the 103 so-called "nebulae" in Messier's catalog 
were in fact clusters of faint stars packed so closely together that the 
individual stars could not be separated or resolved in poorer telescopes. 
Naturally, astronomers speculated whether all the nebulae would even- 
tually be resolved if large enough telescopes could be trained on them. 
For a long time astronomers believed that nebulae were "cosmical 
sandheaps too remote to be resolved into stars." In particular, Bevis' 
nebula, mi, always gave the impression when seen in better and better 
telescopes that it was just on the point of being resolved. William 



Herschel inspected it many times in the course of what he called his 
"star-gaging" with the vast telescopes which he had made specially for 
sweeping the entire heavens. In 1818 he wrote, "it is resolvable" (but 
not, notice, "it is resolved") and went on: "There does not seem to be 
any milky nebulosity mixed with what I take to be small lucid points. 
As all the observers agree to call this object resolvable, it is probably a 
cluster of stars at no very great distance beyond my telescopes' gaging 
powers." His son, John Herschel, placed mi first in a sequence of 
nebulae which had turned out to be clusters of stars, resolvable by 
successive degrees, presumably because he too could glimpse the 
"lucid points" noted by his father. In fact, mi was said by John Herschel 
to be "hairy" or "filamentous." 

The nebula was observed by the Earl of Rosse with his telescope of 
three-foot aperture. In 1844, writing on his observations of nebulae, 
Rosse wrote: "Now, as has always been the case, an increase of instru- 
mental power had added to the number of clusters (of stars) at the 
expense of the nebulae, properly so called ; still it would be very unsafe 
to conclude that such will always be the case." But in spite of Rosse's 
caution here, he too was of the opinion that mi was a star cluster just 
beyond the resolution of his telescope as it had been to Herschel's 
smaller instruments: "It is studded with stars mixed, however, with a 
nebulosity probably consisting of stars too minute to be recognized." 

Rosse added a novel aspect to the description of mi, calling it for the 
first time the Crab Nebula. He wrote that, with his telescope, "it is 
transformed to a closely-crowded cluster, with branches streaming off 
from the oval boundary, like claws, so as to give it an appearance that in a 
measure justifies the name by which it is distinguished." Rosse pub- 
lished a very crustacean-like picture of mi in 1844 and it has been known 
as the Crab Nebula ever since. 

Rosse created a memorable name for the nebula, but it is sad to say 
that its actual appearance does not live up to its imaginative title. His 
contemporary, Dreyer, said that Rosse's 1844 drawing was not at all like 
the real nebula, and 20 years later Rosse published a completely 




British aristocrat Lord Rosses first published picture of the Crab 
Nebula christened it with its distinctive name, but this representa- 
tion looks nothing like Rosse's subsequent accurate drawings of the 
real nebula. 

different drawing, made with his six-foot telescope, then the world's 
largest, which he erected near Birr Castle in Ireland. Slung between two 
towers, the telescope could move only up and down and had very limited 
ability to track stars as they crossed the north-south line through the 
telescope. With it, however, Rosse was able to show that many nebulae 
had a characteristic spiral shape, and we now know that these are 
distant galaxies mostly too far away for the individual stars to be seen. 
Although observing with his telescope must have been difficult, the fact 
that he was able to give such accurate impressions of spiral galaxies 
shows that what he and his co-workers saw with his larger telescope 
and what he drew had firm roots in reality. His later picture of Mi is in 
fact close to modern photographs. 

The reason for his more fanciful earlier picture is not clear but others 
followed him in perceiving something strange about mi. William Lassell, 
an English amateur astronomer who was a brewer by profession, also 
remarked on the filaments and "claws" that he could see when he 
viewed mi in 1853 from the clear skies of Malta. Lassell astutely noticed 


Rosse drew this representation of the 
Crab Nebula from observations with 
his six-foot telescope, and measured the 
positions of the stars he labeled with 
Greek letters. The bay to the east {left) 
and extension to the southwest (lower 
right) can be as clearly seen on this, as 
they can in Roberts'" photograph (below). 


that he could see no more stars in it than were contained in an equal area 
of other parts of the sky nearby, from which he inferred that the stars he 
could see within mi apparently had no connection with it. This marks 
the beginning of the evidence that the Crab Nebula was not going to be 
resolved into stars when a suitably large telescope was built, but would 
remain cloudy and nebular. 

The Crab Nebula was first photographed in 1892 by Isaac Roberts, 
a leading astronomical photographer. Roberts was rather unfairly 
dismissive about Rosse's 1844 descriptions and LasselPs 1853 drawings, 

The first photograph of the Crab Nebula, taken by Isaac Roberts, 
clearly recorded the deep bay at the east (left) side and smaller 
bays to the north (top) and west. Mottling due to the filaments 
was reported to be visible on the original negative. 



to which he said his photograph showed no resemhlance, though he 
acknowledged that Rosse's picture of 1861 had several features that 
approximately corresponded with the photograph. He described the 
nebula as elongated, irregular in outline with a deep bay on one side 
counterbalanced by a projection on the other. The original negative, he 
said, showed mottling, rifts, and some starlike condensations in the 
nebulosity. Apparently Roberts was being careful in saying "starlike" 
here; he seems unwilling to say that the condensations are stars, and his 
caution has since been justified. 

So how does one discover whether the Crab Nebula is made of stars 
or gas? The first major step forward came when two astronomers at 
Harvard Observatory, Joseph Winlock and E. C. Pickering, studied the 
nebula through a device called a spectroscope. 

The spectroscope ranks next to the telescope itself in its importance 
to astronomers, since it has the power to analyze the light from glowing 
bodies. It is worth spending some time explaining how it does this, and 
what the results mean. 

It was Isaac Newton who first found that white light passed through 
a triangular glass prism is split into the colors of the rainbow— what 
scientists call the spectrum. In the early spectroscopes, light gathered by 
an astronomical telescope was split up by prisms and then examined 
visually with a small telescope. 

In the first applications of the spectroscope to astronomy it was found 
that the spectrum of light was continuous — it spread into all the colors 
of the rainbow. A rainbow is in fact the spectrum of a star — the Sun — 
spread into its constituent colors by the prism-like action of raindrops. 
But William Huggins found in 1864 that a large number of nebulae 
showed not a continuous spectrum, but one in which only a small 
number of individual colors called spectral lines occurred. The explana- 
tion for this is that light from a star represents energy given out by the 
particles of which it is made. A single color represents a packet of energy 
of a certain size. If the object on which a spectroscope is focused is 
giving out packets of energy of all conceivable sizes, its spectrum 



contains all colors. This is the case for the light from the dense parts of a 
star's atmosphere. 

The star is composed, like the Earth, of individual particles of matter 
called atoms which are the basic building blocks of the chemical ele- 
ments. Unlike atoms on Earth, however, the atoms in a star are hot 
enough to be moving rapidly back and forth, colliding with each other 
violently, and the force of the collision is enough to fragment the atoms 
into some of their constituent pieces. The outer bits of the atoms can be 
knocked free and these bits join in the jostling of the hot atoms. The bits 
are called electrons. Unless it has been fragmented each atom has a 
precise number of electrons, which makes it the kind of atom that it is. 
The central part of an atom, its nucleus, is positively charged, and 
electrons are negatively charged ; therefore the atom is tied together by 
electrical forces. Only the violent collisions in hot gases at temperatures 
like those found in stars can overcome the electrical forces. The result 
is that a star's atmosphere consists of a dense gas of electrical particles, 
including electrons, rushing back and forth, jostling each other. 

Now light of any color (and indeed radio waves and X-rays as well) 
is commonly produced by just one basic process, though that process can 
occur in a range of circumstances. Whenever a charged particle is 
decelerated (or the direction of its motion is changed, which amounts 
to the same thing), it radiates a pulse of energy, called a photon. The 
wavelength (color) of the energy depends partly on the violence of the 
change in motion. So in a star, the random jerks experienced by the 
electrons produce a whole range of wavelengths, and we see a con- 
tinuous spectrum. 

It is worth mentioning in passing that this continuous spectrum 
usually has a predominant color, depending on the temperature of the 
star. As objects get hotter they glow first red, then yellow, then white, 
then blue as their peak of energy shifts farther to the energetic violet 
end of the spectrum. The electrons in a hot star generally move faster 
than in a cool star, and the cool star therefore appears redder than the 
hot star. 



In the same way, the crowd at a football match have more violent 
encounters with one another than an equally densely-packed crowd at, 
say, a chess tournament, even though a few individual members of 
either may be similarly excited or downcast. So there is a spread of 
energies within both crowds, just as there is a spread of colors emitted 
by stars of different temperatures, but one degree of excitedness 

Suppose, having looked with a spectroscope at a dense gas such as 
exists in stars, we turn to inspect a more rarefied gas. This means that 
there are long distances between atoms, and encounters are few and far 
between. Little continuous light is given out. The light from such a gas 
is not the energy emitted from random encounters but from energy 
changes which take place among the electrons inside the atoms them- 
selves. In a rarefied gas, the atoms have more chance of retaining their 
own electrons, whereas in the chaotic conditions within a star, the 
electrons are usually free to move around. All atoms of a given kind 
contain precisely the same energy levels, like a series of steps, and as an 
electron moves from one level to the next it gives out a step of energy 
of the same size as an electron undergoing the same step in any other, 
similar atom. All atoms undergoing this energy change give out light of 
the same color — a spectral line. 

Thus a dense material gives out a continuous spectrum but only a 
rarefied gas can give out spectral lines. Although the full explanation 
was not known to Huggins, his own observations showed that since 
obvious star clusters always had continuous spectra, while unmistakably 
gaseous nebulae always showed bright individual spectral lines, the 
spectroscope offered a ready means of deciding uncertain cases. Many 
showed a green spectral line, unknown before its appearance in nebulae, 
and thought possibly to come from a kind of atom not present on Earth. 

Indeed, a few years previously astronomers had found several strong 
lines in the Sun's spectrum which did not tally with any gas then known 
on Earth. The mystery gas was christened after helios, the Greek for 
"Sun," which is how helium got its name. Similarly, the new element 



in the nebulae was named nebulium. Later analysis showed that the 
green line was actually the oxygen atom in a state unfamiliar on Earth. 

The light from the Crab 

Huggins does not seem to have observed the Crab Nebula 
through his spectroscope but Winlock and Pickering at Harvard in 1868 
saw the green nebulium spectral line, which proved that at least part of 
the Crab Nebula was gaseous. 

They reported however that the spectrum showed also an unusually 
strong continuous component. This must have been puzzling as it 
suggested that there were more stars than usual embedded in the 
nebula. In fact, such an interpretation is wrong. The continuous spec- 
trum of the Crab Nebula is caused in a quite different way from that of 
a star— it arises from the interaction between electrons and a strong 
magnetic field. This pale clue to a very important process in astrophysics 
was first noticed in particle accelerators called synchrotrons— hence its 
name of synchrotron radiation. 

Modern color photographs show that the two spectral components of 
the Crab Nebula come from distinctly different structures. The nebula 
consists of a lace of red and green filaments, generally oval in total 
outline, embedded in and around a tenuous, milky-white light. Many of 
the filaments are green where the green "nebulium" spectral line 
predominates butfilaments of a red color are common. The red spectral 
line comes from electrons dropping down the first energy step in atoms 
of hydrogen, and is called H-alpha. It was not seen visually, probably 
on account of its deep red color which occurs where the eye's color 
sensitivity is poor. Further, weaker spectral lines from other energy 
steps in the hydrogen atom, called H-beta, H-gamma and so on, are 
present in the spectrum of the filaments as well, shining green and blue. 
Other spectral lines from sulfur, helium and neon can be seen in a 

Knowing that the nebula has this curious knotted appearance of 



Two faces of the Crab. Mt. Wilson astronomers filtered the light of the 
ioo-inch telescope to restrict the wavelengths used in making this pair of 
photographs. Two components of the CraFs light are seen: blue light {left) shows 
a featureless glow while red light {right) picks out the lacy filaments of 
hydrogen gas. 

filaments embedded in a smoother uncolored component, we can see 
now why the 18th and 19th century visual observers always had the 
impression that the Crab Nebula was about to be resolved into stars. 
Descriptions such as John Herschel's "hairy" referred to the filaments, 
while the knots fooled Rosse into believing that he had really glimpsed 
stars embedded in nebulosity. If he had called the stars "lucid points" 
or "starlike" (as did William Herschel and Isaac Roberts respectively) 
he would have given the most accurate visual description of all. 

Motions in the Crab Nebula 

The mere existence of spectral lines in the spectrum of the 
Crab Nebula gave astronomers the vital clue that they needed to decide 
that it was made not of stars but gas. Measurement of the precise wave- 



lengths of the lines would enable them to study the motions of the gas 
in the Crab. Nineteenth century astronomers understood the principle 
of how this would be accomplished but could not marry theory with 
practice until the invention of the photographic plate and its application 
to astronomical spectroscopy. The principle they hoped to exploit was 
the "Doppler effect." 

Doppler shifted violinists 

The Doppler effect is a shift of star's spectral line because 
of the star's motion towards or away from Earth. It was first explained 
in 1842 by Christian Doppler, an Austrian physicist, with reference to 
sound waves rather than light waves. For one experiment, he hired a 
group of violinists, who had the gift of absolute pitch so that they could 
tell precisely what note an instrument was playing without reference to 
any external standard. Doppler asked the musicians to sit in an open 
railroad car and play one particular note as the car moved at various 
speeds along a length of track. While they did this, another musician 
stood by the track and listened to the note that the violinists were playing. 
When the car was approaching the stationary violinist, he heard a higher 
note than the one actually being played. When the car was receding, 
he heard a lower note. Most people, standing at a railroad station, have 
experienced the difference in sound between an approaching train and 
one that is receding. 

There is close correspondence between what happens in the case of 
sound waves, and what happens in the case of light waves. We talk 
about colors or spectral lines instead of musical notes but the two con- 
cepts are identical. Instead of a "higher note," we talk about a "blue- 
shifted spectral line," and instead of a "lower note," a "red-shifted 
spectral line." The degree to which a spectral line is moved towards the 
blue or red end of the spectrum is a measure of how fast the object which 
emitted this light is moving towards or away from us. No matter how 
distant the star, as long as spectral features can be distinguished 



shorter here 
train approacnin' 
higher note he 

longer here as 
train receding — 
lower note heard 
by person with 
absolute pitch 

In C. Doppler's experiment successive waves of sound were emitted by the 
violinist as he moved along the track on the flatbed car. The waves were 
crowded together in the direction of motion of the train, spread apart to rear of 

astronomers can measure its speed along the line of sight. If the star is 
fairly close, it will also change its position in the sky over a period of 
time and so its motion in all directions is known. 

In practice the Doppler shift in a typical astronomical object is small. 
The wavelength of its spectral lines might change, typically, by 0.03 7 . 
Astronomers' first attempts to measure Doppler shifts were foiled by 
their smallness. A spectrum had to be examined in fine detail before 
they could be perceived. Of course the finer the detail in which you need 
to look at anything the more light you need to be able to see it, just as to 
read small print you need a brighter light than to read a newspaper 
headline. The eye could not grasp enough light to "read" fine Doppler 
shifts in the first spectroscopes. Only when the astronomer's eye was 
replaced by a photographic plate to record the spectrum was it possible 



to measure the Doppler shift caused by motions of stars and nebulae. 

The photographic plate can do what the eye alone cannot : it integrates. 
That is to say, unlike the eye which perceives a new image about every 
-it second, a photographic plate can be put at the focus of a spectroscope 
for many hours to store up, or integrate, the photons that the sky is 
sending. Since the spectrum appeared as a permanent picture rather 
than a fleeting image in a man's eye, the spectroscope was renamed 
the "spectrograph." With a spectrograph it became possible to record 
the wavelengths of light and the way in which light was distributed 
among the various wavelengths in finer and finer detail. 




A stationary nebula gives single emission lines in its spectrum, but the 
expanding Crab nebula showed doubled lines in VM Slipher's spectograph 

EARTH r ; . -v:' 

This side 

approaching Earth 
— gives blue 
shifted line 

This side 

receding from Earth 
— gives red 
shifted line 



blue shifted 


■—♦► red 



In 1913 V. M. Slipher turned his spectrograph towards the Crab 
Nebula and photographed its spectrum. Whereas visual observers, using 
low-powered visual spectroscopes, had seen individual spectral lines, 
on his photographic plate he was able to distinguish that every spectral 
line was doubled: it appeared twice, once shifted a small amount 
towards the blue end of the spectrum and once to the red. He imme- 
diately recognized that this was a manifestation of the Doppler effect 
and that it meant that the light from the Crab Nebula came from two 
parts, one of which was receding from us and one of which was moving 
towards us. He correctly deduced that this was because the Crab Nebula 

l 1- 

This extraordinary photograph is of the far side of the Crab 
Nebula. It was taken by Royal Greenwich Observatory astronomers 
D. McMullan, P. Wehinger and R. Fosbury using a colored filter 
which selected only light from the redder Doppler-shifted part 
of the nebulium spectral line from the receding side of the Crab. 
Light from the approaching side whose nebulium line is Doppler- 
shifted to the blue was excluded. 



'j^z&tpr ' ■'■■♦ ■■■ 

■ ;X ,;s -i i * : '"' . ■ '■- - v' ■ 

w- ^sT 

7^ expansion of the Crab Nebula over a period of 14 years is shown 
graphically in this photograph made by Virginia Trimble. This is a double 
exposure of a pair of prints taken with the 200-inch Mt. Palomar telescope in 
ig$o and 1964. The stars scattered over the photographs were superimposed 
precisely so any movement in the intervening years would be clearly seen. 
Trimble was thus able to confirm that the filaments were close together at a 
central point some 800 or goo years ago. 

was expanding, with the nearer side approaching us and the farther side 
going away, at speeds up to 1000 km/s. 

Slipher's observations of the Crab Nebula's spectrum were repeated 
by R. F. Sanford at the Mount Wilson Observatory in 191 9. Then in 
1 92 1, C. O. Lampland photographed the Crab Nebula with the 40-inch 
Lowell reflector and compared that photograph with an earlier one. He 



was able to see that changes had taken place in the Crab Nebula. This 
provoked speculation that the changes in appearance were a consequence 
of the high expansion speeds but Lampland could not distinguish 
between changes in brightness in individual parts of the nebula and 
motions caused by expansion, because his pictures had been taken on too 
small a scale. 

Then John Duncan photographed the nebula with the 6o-inch 
reflector at Mount Wilson and compared his photograph with an excellent 
one made in 1919 by G. W. Ritchey with the same telescope. When 
Duncan compared the two photographs, he was able to see the changes 
that Lampland had announced, and he was able to tell from his larger 
scale photographs that many of the outer filaments and parts of the 
nebula had unmistakably moved outwards. Thus the nebula was actually 
seen to be expanding, and the spectrographic results were clearly 

Duncan published his result in 1921, and in the same year, by a 
remarkable coincidence, the Swedish astronomer K. Lundmark pub- 
lished a list of novae that had been observed by the Chinese. Number 36 
in his list was the supernova of 1054. Lundmark noted that the position 
of that supernova was very near to mi, the Crab Nebula. 

Identifying the Crab 

None of the other astronomers who had investigated the 
Crab Nebula made the necessary mental jump of connecting the 
supernova of 1054 with the expansion of the nebula. This was left to one 
of the most famous names in astronomy, Edwin Hubble, in 1928. 
Hubble's name is usually associated with objects outside the Galaxy — he 
first measured the expansion of the Universe, and the "Hubble Time" 
defines the age of the Universe itself. He was one of the leading observers 
of his day, and investigated many nearer astronomical objects, including 
the Crab. By comparing the size of the Crab with its observed rate of 
expansion, Hubble was able to estimate that some eight or nine hundred 
years had elapsed since the expansion began. 



. ■. : . '-■..-' 


- - ' '■.■■ 

'■"■". -i ■ ; . /■■' «■"'" '■■■.' 

Photo {left) of the central area of the Crab Nebula was taken about the time of 
the abrupt period decrease of the Crab pulsar m September ig68. By the time 
photo (right) was exposed, 124 days later, the Thin Wisp had moved out 
towards the wisps and Wisp 1 had become separated from the others. These Lick 
Observatory photos (shown as negatives) taken by J. Scargle confirmed Baade's 
earlier suspicion that "waves of activity'" spread into Crab Nebula from the 
stellar remains of the supernova of 1054, and linked their cause to the Crab 
pulsar spin-ups. 

Hubble was the first to point out the coincidence of position and age 
between the Crab Nebula and the supernova of 1054. He published his 
inspired guess in a series of popular essays on astronomy and it escaped 
the attention of professional astronomers. It was not until two Dutch 
scholars, one an astronomer and one an orientalist, worked on the prob- 
lem during World War II, that the identification became accepted. 

The astronomer was Jan Oort, a leading contributor to knowledge 
about our Galaxy, and the orientalist was J. J. L. Duyvcndak. Working 
in Holland under the German occupation, they sent their discoveries via 
Sweden, a neutral country, to N. U. May all in the United States. 



The central area of the Crab Nebula is shown in this photo taken 
in ig68 by J. Scargle with the Lick 120-inch telescope. In this 
negative print, brighter objects appear blacker. The pulsar is 
the lower right (southwest) of the two central bright stars, touched 
by the Thin Wisp. The bright nebulous area to the upper right is 
subdivided into two of the three Wisps (the third was not visible 
at the time this photo was taken). Directly opposite the area of 
the Wisps is the Anvil, also subdivided into two wisps on this 

Mayall published the Chinese descriptions of the supernova, which were 
discovered by Duyvendak, in an astronomical discussion in 1942 — a 
remarkable example of the way in which science transcends political 
divisions, even during wartime. A few years previously, at the Lick 
Observatory, Mayall had obtained excellent spectra of the Crab Nebula, 
showing very clearly its expansion as evidenced by the Doppler effect. 
He found that the speed of expansion was 1300 kilometers every second. 
Expanding at this speed for eight or nine hundred years, the Crab 
Nebula had become about seven light years in diameter. 

Knowing the true size of the nebula, astronomers could then calculate 



just how far away it is. Since this method is one often used by astrono- 
mers to determine distance, it is worth explaining exactly how it operates. 
Consider some object with a fixed size, say a ball i foot in diameter. 
At a distance of, say, 10 feet the ball has a certain apparent size. At twice 
the distance its apparent size has diminished to half. It appears smaller 
because of its distance although we know that it still is i foot in diameter. 
Clearly, then, there is a relationship between the distance of an astro- 
nomical object, its apparent size and its real size, and if astronomers 
know any two of these quantities they can calculate the third. In the case 
of the Crab Nebula, Mayall determined its true size by multiplying the 


Wisp 3 

Pulsar here 
in 1054—, 

of Pulsar 

Center of nebula's - 
expansion measured ', 
to lie somewhere 
in this circle 

Waves of activity spread from the southwest star SW into the 
Thin Wisp, The Anvil and the three Wisps 1-3. The SW star 
was at X in 1054, within the circle which contains the center of 
the Crab Nebula's expansion. These two facts pointed out the SW 
star as the active star in the Crab Nebula before it was determined 
to be the pulsar. 



expansion speed by its age, and measured its apparent size in angular 
degrees, and so could find its distance. 

The best modern determination of the distance of the Crab Nebula is 
5500 light years, meaning that light takes that length of time to travel 
from the nebula to Earth. The supernova which the Chinese observed 
in 1054 had actually exploded in around 4500 bc, and the light from the 
explosion had been traveling for about 5500 years to reach the Earth on 
July 4, 1054. 

Measurements of the expansion of the Crab Nebula tell astronomers 
not only when the expansion began, but also the place where it began. 
The filaments had exploded outward and the place from which they 
radiated is the place where the explosion occurred. The center of the 
explosion lies near two stars at the middle of the nebula. 

Because of their positions in the sky they became known as the 
southwest star and the northeast star. The expansion center is actually 
now equidistant from both stars. However, the southwest star has a 
so-called proper motion of its own. (The word "proper" is not used in 
any sense of moral rectitude, but because the motion belongs to the star 
itself and has nothing to do with the motion of the Earth, as, say, does 
its rising and setting.) The northeast star has no discernible proper 
motion. In their 1942 paper on the Crab Nebula, Walter Baade and 
Rudolph Minkowski took into account the motion of the southwest star 
since 1054 and found that this star was closer to the original site of the 
explosion. The northeast star is probably not in the nebula at all, but 
lies in that direction by chance. 

The southwest star lies in a hole in the center of the Crab Nebula, 
surrounded by little wisps, bays and filaments which seem to change in 
brightness and position. The changes were among those first seen by 
C. O. Lampland in 192 1. Walter Baade studied the changes in the center 
in detail and found in 1945 that the brightness changes gave the im- 
pression that light was rippling in waves outwards from the southwest 
star. He took photographs with the 200-inch telescope to prove his point, 
but had not completed analyzing them when he died in i960. Guido 



Munch and Jeffrey Scargle subsequently used these plates and later ones 
they took themselves to confirm Baade's work. 

Baade and Minkowski strongly suspected in their 1942 paper that the 
southwest star was not only the central remnant of the supernova 
explosion, but that it was still affecting the central regions of the 
nebula. They were correct. The southwest star is in fact the Crab 
Nebular pulsar, as we shall see in the next chapter, and waves of activity 
propagate from this star throughout the nebula. 

The accelerating Crab 

The most accurate estimates of the Crab's expansion 
actually put the date at which it began to within 10 years of ad 1140, 
significantly different from 1054. Has there been some strange mistake? 
In fact, astronomers interpret this as evidence for another of the Crab 
Nebula's peculiarities, and one which once fooled Baade himself. 

The estimate of the date when the expansion began is based on the 
assumption that the nebula has been expanding throughout its history 
just exactly as fast as over the last 30 years (the time interval between the 
photographs used to measure the expansion speed). The identification 
of the Crab with the 1054 supernova is so strong that this assumption 
has to be wrong. But, rather surprisingly, the nebula is not slowing 
down : it is speeding up. If it were slowing its present expansion speed 
would now be smaller than in the past and we would mistakenly say that 
the explosion occurred earlier than 1054. Since it appears from the 
expansion rate that the explosion occurred a hundred years later than 
1054, the present speeds are larger on average than the speeds have been 
over the last nine hundred years. 

This was first discovered by Walter Baade, using Duncan's measure- 
ments, but at first he thought that the observations were not accurate 
enough. He simply could not believe the result because he would have 
expected the expansion to slow down as the nebula crashed into the 
surrounding interstellar material. But it does seem to be the case that 



the nebula is accelerating. What that implies is that some extra energy 
is available from somewhere to drive the expansion at faster and faster 
speeds. The initial energy of the supernova explosion has caused it to 
start expanding but some additional energy has been pumped into the 
nebula to make the expansion faster. 

Before the mystery of the source of the energy accelerating the 
filaments could be solved, radio astronomers found further proof that 
there was an active powerhouse in the Crab, and that it wasn't just using 
up its inheritance of energy from the supernova explosion itself. 

The Crab among the radio stars 

Astronomy, and particularly radio astronomy, is full of 
unexpected discoveries, which pop up through serendipity, as the result 
of looking for something else. Indeed, the man whose name is now 
indelibly linked with beginning radio astronomy did not set out to do 
anything of the sort. The man was an engineer at the Bell Telephone 
Laboratories, and his name was Karl Jansky. 

Bell were interested in the hiss which represented the ultimate limit 
to the sensitivity of radio reception and transmission, and therefore set 
Jansky to investigate atmospheric interference. To do this, he built an 
antenna which he called the Merry-go-Round, because it could be 
rotated to track down the source of the hiss. In December 1931, Jansky 
noticed a source of radio interference whose intensity cycled with a 
period of 23 hours 56 minutes, that is, it was keeping pace with the 
rotation of the Earth with respect to the stars and not recurring every 
24 hours which is the period of rotation of the Earth with respect to the 
Sun. This led Jansky to conclude that the hiss was of cosmic origin. It 
peaked in the Milky Way constellation Sagittarius. 

Because of commercial pressure, Jansky was unable to continue this 
investigation, and the first radio maps of the Milky Way were made by 
an American amateur astronomer, Grote Reber, in the early 1940s, using 
a home-built backyard antenna. They showed a broad swathe of radio 



noise coming from the Milky Way, most strongly towards the galactic 
center in Sagittarius. There were two subsidiary peaks in the constella- 
tions Cygnus and Cassiopeia blending into the general Milky Way. 

Radio astronomy had been studied during World War II largely 
because natural cosmic radio noise affected radio reception and the 
operation of radar. With the ending of the war, it became possible to 
investigate these phenomena more fully and the opportunity to do so 
was provided by large quantities of surplus wartime radio and radar 
equipment. A radio telescope, after all, is no more than a sensitive radio 
receiver coupled to a directional antenna, such as those used for radar. 

A team in England led by John Hey had found during wartime 
research that the Sun was a source of radio noise when there were large 
sunspots, and that meteors showed on radar. After the war this team 
mapped the intensity of radio waves along the Milky Way and noticed 

A radio telescope's view of the Crab. The basic map has brightness contours 
only, which here have been shaded in. 


the bright pointlike source of radio waves in Cygnus. Hey's radio 
telescope had a narrower beam than Reber's so that the Cygnus source 
stood out as more pointlike, but the blurriness of Hey's telescope's 
beam was still 16 times the area of the Moon: its resolution was 2°. 

Imagine the optical appearance of the night sky seen with a 2° beam. 
In fact you can simulate such a beam by looking through very defocused 
binoculars at the sky. Seen like this, the Moon has no surface detail at all. 
Its phases cannot be distinguished; all that can be seen is its change of 
brightness through the month. The appearance of the Milky Way is 
about the same whether seen through these binoculars or not. The 
blurred shapes of all but the brightest stars merge into the Milky Way 
and are indistinguishable from it. The brightest stars, however, can be 
seen as blurred disks. The view hints at crisper detail beyond the 
resolution of the binoculars. 

The telescope that looked to sea 

Arguing by such an analogy, some radio astronomers put 
effort into determining whether Hey's radio source in Cygnus was truly 
pointlike, and whether there were other radio sources like it. But how 
could they make a radio telescope that pointed well enough? In 
Australia, John Bolton and G. J. Stanley used a radio telescope situated 
on Dover Heights, an eastern suburb of Sydney. They set the telescope 
on a cliff top overlooking the Pacific Ocean and directed it towards the 
radio stars as they rose. It received radio radiation not only directly from 
any radio stars that rose above the horizon, but also from their reflection 
in the sea. In effect, it functioned in some ways like two connected radio 
telescopes, the real one 300 feet above sea level on the cliff top, and 
another one, 300 feet below the sea. Such a radio telescope, called an 
interferometer, has the same ability to discriminate fine detail as a single 
vast radio telescope 600 feet in diameter. 

With this telescope, Bolton confirmed the existence of the Cygnus 
radio source and also saw a radio source in Taurus, which he named 



Taurus A. His method of determining a radio source's position in the 
sky consisted in looking at the time at which it rose above the horizon 
and at the rate at which it ascended above it. However, there was a 
problem analogous to atmospheric refraction. The effect of air on the 
light waves from stars at the horizon is to make them visible before they 
have actually risen above the horizontal. This effect is seen in an 
exaggerated form in a mirage when light from an oasis below the horizon 
can be refracted into the gaze of a desert traveler. Normally, atmospheric 
refraction causes stars and the Sun to rise at least two minutes early and 
set at least two minutes late. Another effect is that changes in the degree 
of refraction cause stars near the horizon to twinkle more than stars 
overhead. Radio refraction in the ionosphere above Earth is similarly 
highly variable and is more severe than optical refraction in the atmos- 
phere. Moreover, when radio refraction is strongest, it is least variable 
and causes least twinkling in the appearance of the radio stars. 

Bolton selected the clearest records of Taurus A rising, and in doing 
so he had selected those with abnormally large radio refraction. Conse- 
quently, the first estimate of the position of Taurus A was a long way in 
error. Compensating for this refraction in later experiments, Bolton 
took his interferometer to New Zealand, where he observed Taurus A 
rise from the eastern side of the island, and set from the western side. 
Any error in time of star-rise was compensated by an equal and opposite 
error in the time of star-set. Bolton then looked to see what optical 
sources were in that part of the sky, using his only suitable reference 
work, a star atlas used predominantly by amateur astronomers, Norton's 
Star Atlas. There he found marked mi, the Crab Nebula. Thus in 1948, 
a radio star was identified with a visible object for the first time. 

Bolton and Stanley later used a genuine twin telescope interferometer 
to observe Taurus A when it was well above the horizon and the effects 
of refraction were small. They tied the position of Taurus A down to 
within 15 arc minutes, an area one quarter of that of the Full Moon. 

A question immediately arose : was the radio emission coming from 
the Crab Nebula itself or from a star embedded in the nebula ? A two- 



pronged attack brought the first answers. The initial approach was by 
various technical advances in making radio interferometers of finer and 
finer resolution, by placing the radio antennae farther and farther apart. 
A Manchester group led by Robert Hanbury Brown worked with 
separations up to 4 km. In Australia, Bernard Mills used two radio 
telescopes up to 10 km apart to obtain a viewing beam about 1 arc 
minute wide. The second line of attack was afforded by a lucky chance. 
Taurus is not only a Milky Way constellation, it is also on the ecliptic, 
the yearly path of the Sun, Moon and planets. The Moon, in fact, 
eclipses the Crab Nebula, and it is possible to see the gradual fading of 
Taurus A as the Moon moves in front of it. Eclipses in 1956 showed that 
the radio emission comes from the whole of the nebula, because it is 
covered gradually as the Moon advances across it. With the technically 
superior interferometers now available at Cambridge, A. S. Wilson has 
made the most detailed maps of the radio emissions from the Crab, 
showing that its shape is remarkably like that of the white light com- 
ponent of the Crab Nebula, with the same broad hump, dissected by the 
bays and valleys that show at optical wavelengths. 

X-rays from the Crab 

In 1963, the Crab Nebula was discovered to be emitting 
X-rays. X-radiation is a form of radiated energy, like light, but a 
thousand times more energetic. When X-rays shine from outside the 
Earth onto atoms in the upper atmosphere, such as oxygen and nitrogen, 
the X-rays are quickly and readily absorbed. In this process, electrons 
orbiting deep within the atoms take up the X-ray energy, causing the 
electrons to be ejected from the atoms. X-rays traveling at sea level are 
typically completely absorbed by air after a few hundred feet or so. 
(Similar X-rays are used in X-ray machines to photograph bones of the 
body: the areas where X-rays have been absorbed, by just a few inches 
of flesh and bone, show up as shadows on a photograph.) 
Because X-rays are so readily absorbed, the only way astronomers 



can detect cosmic X-ray sources is by flying X-ray detectors above most 
of the atmosphere. Balloons can carry X-ray telescopes above a good 
deal of the atmosphere, but by its nature, a balloon, needing air to 
support it, cannot escape completely above the air. Only the most 
penetrative and energetic X-rays can be detected in this way. (Balloons 
also have the astronomical disadvantage that they obscure the most 
interesting and least absorbed part of the sky, immediately overhead.) 

Rockets have the ability to carry instruments completely above the 
atmosphere in long parabolic trajectories, without going into orbit, and 
it was by this means that the Crab Nebula X-ray source was detected in 
1963. A rocket flight, however, lasts for just a few minutes, as long per- 
haps as a total solar eclipse, and costs so much that it occurs about as 


Satellites, launched into orbit around the Earth above the atmosphere, 
last for several years, so that although they are more expensive to send 
up, they are far less expensive per minute of observing time than 
rockets. However, the first satellite launched exclusively for X-ray 
astronomy, Uhuru, was not launched until 1971. 

The X-ray telescope with which a group of X-ray astronomers at the 
Naval Research Laboratory, Washington, D.C., discovered the Crab 
Nebula in 1963, accepted X-rays from an area of the sky 20 wide. The 
year before, a group at the American Science and Engineering company 
had flown an experiment which detected an X-ray emitting region in the 
Milky "Way in Taurus, and the NRL rocket was commanded to sweep 
over this area. As the NRL X-ray telescope slewed over the Crab Nebula, 
it recorded a maximum in the number of X-ray counts, and the scientists 
found that the peak was somewhere within 2° of the Crab. In this way, 
the Crab became the first X-ray source to be identified with a known 
celestial object (excepting the Sun), just as it had been the first identified 
cosmic radio source. 

By 1969, the Crab X-ray source had been the subject of more than 
30 rocket and balloon flights carrying X-ray equipment. It is interesting 
to estimate the cost of the effort put into studying the Crab's X-ray 



photons, assuming that an X-ray experiment costs something on the 
order of $100,000 and collects perhaps 10,000 to 100,000 X-rays from 
the Crab. 

The Moon pinpoints X-rays 

The NRL group followed up its work with a further rocket, 
launched so that it would be looking at the Crab during its eclipse by 
the Moon on July 7, 1964. Three minutes after launch, the Crab was 
seen beginning to disappear as the Moon passed in front of it, and it 
slowly faded away to nothing over the next three minutes. During those 
three minutes, the edge of the Moon had scythed across two arc minutes 
of the sky, so that the NRL group had shown that the X-ray emitting 
region was approximately the same size as the optical and radio Crab 

The "seasons" during which the Moon eclipses the Crab recur every 
10 years. A group of X-ray astronomers at the Lawrence Livermore 
Laboratory of the University of California observed the next lunar 
eclipse of the Crab in 1974. They were able to tell that the lower-energy 
X-rays in fact came from an area somewhat smaller than the optical 
nebula, and centered somewhat west of the star suspected as the center 
of the Crab, and near the brightest of the wisps, which Baade and later 

Following a spiral path around a line of magnetic force, an electron in the Crab 
Nebula radiates radio, optical and X-radiation by the synchrotron process. 


T7 ^^^ 



Scargle had found to shift position and change in brightness. Another 
experiment, this time aboard a balloon operated from MIT, showed the 
very high energy X-rays to come from an even more compact area 
centered on the wisps. The X-radiation therefore has a somewhat more 
intimate connection with Baade's southwest star and its immediate 
surroundings than optical and radio radiation, but if unearthly beings 
with strange eyes were to study only the radio or X-ray Crab Nebula, 
they would picture it in much the same way that we do: the invisible 
Crab mimics the visible one. 

What makes the Crab shine ? 

The Crab Nebula looks similar in light, radio waves and 
X-radiation, unlike most other nebulae. And, indeed, all the radio, 
X-ray and the optical emissions are caused by the same phenomenon, 
known as synchrotron radiation. When the Russian astrophysicist 
Iosif Shklovsky proposed this explanation in 1953, the theory was at 
first thought to be absurd. The model it proposed was of a Crab Nebula 
full of fast electrons gyrating about strong magnetic fields and radiating 
their energy over a wide spectrum stretching from radio through optical. 
Shklovsky predicted that the radiation would be polarized so that when 
the optical radiation was photographed through a polarizing filter, the 
nebula would appear streaky along the lines of the magnetic forces which 
pervade it. This was found to be the case, and the theory of synchrotron 
radiation was vindicated. 

Now, a characteristic of electrons radiating by the synchrotron 
mechanism is that they lose their energy relatively quickly. The lifetime 
of a fast moving electron in the Crab Nebula (that is, the time in which 
it radiates half its energy) is much less than the 900 year age of the 
nebula. Therefore, the fast electrons cannot have been spiraling about 
the nebula since the original supernova explosion: they must have been 
injected into the nebula since then. The existence of the synchrotron 
radiation demands that energy must be pumping into the nebula in the 



sV -lll'lij. 

Fritz Zwicky described the polarization pattern of the Crab Nebula as a 
"basket weave." It is well shown in this photographic representation by 
F. Gieseking of the Bonn University Observatory. Polarized light comes from 
electrons gyrating in the Crab's magnetic field which runs perpendicular to the 
lines on this diagram. 

form of fast electrons at this moment. Remember that, from the rate of 
the Crab's expansion, there is other evidence that some powerhouse is 
still functioning in the Crab. Until the discovery of pulsars in 1968 the 
source of this extra energy was a mystery. 


VI Discovering pulsars: 

heartbeats of supernova 


: story of the discovery of pulsars in 1967 is a classic 
among the many scientific tales of perspiration, inspiration, and just 
plain luck. Pulsars linked the practical world of the observing radio 
astronomer with that of the theoretician, who for years had been talking 
about mysterious objects called "neutron stars." And, in explaining 
how they pulsed and how they emitted radio waves, astronomers found 
that they naturally provided an explanation for the amazing expansion 
and acceleration of the Crab Nebula. 

The discovery had the added drama that, for a time, it seemed to be 
evidence for extraterrestrial life, although by the time the news of the 
first pulsar was published, its discoverers were already quite sure that 
the signals were not artificial. 

Even the instrument with which pulsars were first detected has 
curiosity value in its own right. While most radio telescopes are mea- 
sured by their size in feet, meters, or even miles, this one has a splendidly 
archaic name, which never fails to confuse foreign astronomers: it is the 
4^ acre telescope at Cambridge, England. 

It was designed for a purpose which had nothing to do with the 
regions of space where pulsars lurk. It was, and in its present enlarged 
form still is, designed to pick out those enigmatic radio sources known 
as quasars, by means of their scintillation or twinkling. 



Every amateur astronomer knows that you can easily tell a planet 
from a star because a star, being a point of light, usually twinkles whereas 
a planet doesn't. Twinkling is caused by unsteadiness in the Earth's 
atmosphere. A planet twinkles much less, if at all, because in a telescope 
it has a discernible disk, even though to the eye it may seem the same 

size as a star. 

Much the same sort of thing happens with radio sources, though 
radio astronomers use the word "scintillation" instead of "twinkling." 
The medium causing the disturbance is thin gas called plasma. Not only 
is plasma found in the atmosphere of the Earth but also far out in the 
solar system, between the planets. 

For Antony Hewish and his team, who first picked up interplanetary 
scintillation from pointlike radio sources in 1964, its importance was 
that it enabled the tiny, distant quasars (quasi-stellar radio sources) to 
be picked out from nearer, apparently larger, sources. 

To get the sensitivity necessary to distinguish rapid fluctuations of 
signal, Hewish needed a radio telescope with a large collecting area. 
This he achieved by simply setting up wires on poles, covering a paddock 
of 4^ acres. This had the required ability to pull in faint signals, but it 
lacked the direction-finding discrimination of a dish-type antenna. By 
July 1967, the new telescope was ready to begin recording. It was 
designed to scan a large part of the sky in one week. The equipment was 
built to emphasize the scintillation rather than, as was usual, to de- 
emphasize it, and it was able to respond in as short a time as one tenth 
of a second to fluctuations in the brightness of a radio source. Hewish 
wanted all the radio sources which it found to be plotted on a map of 
the sky, so that terrestrial manmade interference, which would appear 
randomly, could be sorted out from the truly extraterrestrial twinkling 
radio sources which would recur at the same celestial coordinates. The 
person to whom he assigned the job of analyzing the data from the 
instrument was a graduate student, Jocelyn Bell. 


Scruffy little green men ? 

Analyzing the data from the new telescope was no small 
task. The instrument produced 400 feet of tape from each scan of the 
sky, 100 feet every day. Bell's job was to examine every signal, discarding 
such manmade phenomena as aircraft transmissions and foreign tele- 
vision stations and mapping out the true extraterrestrial signals. By 
October, she was 1000 feet behind current chart production and yet, 
fortunately, she did not relax her standard of attentiveness. 

It was in October that Bell noticed what she called "a bit of scruff." 
It was passing through the beam near midnight when the interplanetary 
scintillation normally falls to a low level, because at this time the radio 
telescope is pointing to the outer edge of the solar system where the 
plasma is least dense. Bell's account says, " Sometimes within the record 
there were signals that I could not quite classify. They weren't either 
twinkling or manmade interference. I began to remember that I had 
seen this particular bit of scruff before, and from the same part of the 

The source seemed to be recurring every 23 hours 56 minutes, and 
only objects fixed among the stars recur every 23 hours 56 minutes. 
Manmade interference, though, tends to recur on a 24 hour schedule, 
because daily life is ordered by the Sun. The moment when Bell 
recognized that the bit of "scruff" was more than a single piece of 
interference, but had actually occurred before at the same celestial 
coordinates, proved to be very important. Bell describes her reaction: 
"When it clicked that I had seen it before I did a double take. I remem- 
bered that I had seen it from the same part of the sky before." 

Looking back at the records, Bell was able to prove that she had, in 
fact, seen it two months before, in August. She then discussed the signals 
with Hewish. They decided to use the observatory's fast recorder to get 
a clearer picture of the nature of the signals. When the fast recorder 
became available in mid November, Bell was given the job of trying to 
catch the signals and record them. For some days she was unsuccessful. 



1 1 1 1 1 



| | , t 1 1 

■ ■ 

; ^ HI- 

l.l. 1 1 

O 10 20 30 40 

How radio astronomers see a pulsar. This is a recording of the signals from radio 
pulsar oj2g + 54, observed with the National Radio Observatory's ^2-meter 
transit telescope by Richard Manchester. Each spike on the chart is a pulsar 
flash recurring at 0.714 second intervals. 

At this point, Hewish thought that the signals were from a randomly 
occurring flare star and that it was unlikely that they would see it again. 
Bell persevered, however, and at last managed to catch a satisfactory 
recording which showed clearly that the "scruff" was a burst of pulses 
almost exactly i£ seconds apart, similar to many kinds of terrestrial 
interference. When she telephoned to Hewish to tell him what she had 
found, he said, "Oh, that settles it. It must be manmade." 

Nonetheless, Bell and Hewish continued to make recordings of the 
"scruff." The main problem was still that the pulses were keeping 
sidereal time, recurring with the 23 hour 56 minute period. Were the 
bursts truly sidereal, or were they being made artificially with a sidereal 
period ? The only people on Earth who could conceivably be imitating 
sidereal time would be astronomers, though no one could guess why 
they would want to make bursts of pulses like this. Inquiries at other 
observatories failed to reveal any program which could account for the 
signals. Searching around for sidereal explanations for the pulsing signal, 
Hewish and Bell considered whether known variable stars could cause it. 
The trouble was that the fastest variable star known had a period of 
about a third of a day. How could a star throb with a period of 1.337 
seconds ? 



Caught in the dilemma that the pulses were extraterrestrial, but 
seemed to be artificial, the Cambridge astronomers began to consider 
a new possibility, were the pulses being manufactured in space by an 
extraterrestrial civilization ? By mid December, they had proved that the 
pulses recurred very regularly indeed, staying on schedule to one millionth 
of a second. In a half-joking way, Bell's colleagues began to refer to the 
star as lgm-i. But why would Little Green Men manufacture and 
broadcast repetitive signals like this ? Most radio signals change, in order 
to convey information; the constant ones are navigational aids like the 
lor an signals (for LOng RAnge Navigation). Was lgm-i an inter- 
stellar navigational beacon ? 

If the pulses were being manufactured artificially by an intelligent 
civilization, the manufacturers presumably lived on a planet. If the 
signals were coming from a planet, they would show the effects of a 
Doppler shift. The Doppler shift causes a bunching effect of repetitive 
signals as the transmitting object moves towards the recipient, and a 
spacing out effect as it moves away. The Cambridge astronomers had, 
in fact, already observed a small change in the timing of the signals, 
caused by the motion of the Earth around our Sun. Could they see the 
equivalent effect, at a different period, caused by the transmitter itself 
being on another planet orbiting its own sun ? Bell recorded in her log, 
"We are working on the Doppler shift of the pulses to see whether the 
source is stationary or moving round a sun. There is no 4c (Cambridge 
catalog) source with the same coordinates, nor any other source that we 
know of." In the event, no Doppler shift was seen, other than that 
caused by Earth's motion. The radio source was therefore not on 
another planet. The little green men became less likely. 

The Cambridge radio astronomers used the amount of Doppler shift 
on the period of the pulses to estimate the source's position in the sky 
because, as we have noted, the \\ acre telescope lacked direction-finding 
precision. Suppose that a pulsar with a period of one second lies in the 
plane of the Earth's orbit. On a certain day of the year, the Earth will be 
traveling directly towards it with its orbital speed of 30 kilometers per 



second. This will cause a decrease of the period by one ten thousandth of 
a second to 0.9999 seconds. Six months later, the Earth on the other side 
of its orbit will be traveling away from the pulsar at 30 kilometers per 
second, and the period will be lengthened to 1.0001 seconds. 

On the other hand, if this imaginary pulsar is in a direction perpen- 
dicular to the Earth's orbit (a direction known as the pole of the ecliptic), 
the Earth will throughout the year be traveling across the line of sight 
to the pulsar, no Doppler shift will be observed and the period will 
always be 1.0000 seconds. 

Such small changes may seem impossible to measure, particularly 
when we see the indistinct signal from a weak pulsar. But we are able 
to take the average of hundreds or thousands of pulses. After 10,000 
pulses of the hypothetical pulsar, each one delayed by 0.0001 seconds, 
the combined delay will be one whole second. Ten thousand seconds is 
not quite three hours, so in a brief space of time it is possible to measure 
the rate of a pulsar to much better than 1 part in 10,000. From the size of 
the Doppler shift of the first pulsar, the Cambridge team estimated its 
position and that it was unmoving, to an accuracy of about two arc 
minutes. Just as Brahe had shown for the supernova of 1572, and Kepler 
for the supernova of 1604, and with the same accuracy, the Cambridge 
astronomers had demonstrated that the pulsar had no discernible parallax 
and must be farther than the edge of the solar system. 

Measuring the pulsar's distance 

The Cambridge team then proceeded to make some 
measurements which were possible with two radio receivers operating 
simultaneously, but tuned to different radio frequencies.* Pulses 

* The frequency and wavelength of radio signals — or any other kind 
of waves — are linked. The frequency is the number of waves passing a 
point in a given time, while the wavelength is the crest-to-crest 
distance of the waves. As the wavelength gets shorter, so the waves get 
more frequent and the frequency is higher. 



observed on the two frequencies arrived at different times, a pulse 
traveling on the longer wavelength radio arriving later than the same 
pulse traveling on the shorter wavelengths. The delay between the two 
signals showed that the radio frequencies traveled through interstellar 
space at different speeds, a phenomenon called dispersion. 

If interstellar space were truly empty, all radio waves would travel at 
the speed of light. But there are free electrons in interstellar space, 
produced from the ionization by starlight of atoms of interstellar gas, 
such as sodium. Starlight passing near a sodium atom can interact with 
the atom, causing it to eject the loosest of its electrons. Interstellar 
space is thus not empty: it contains a plasma, a low density gas con- 
taining unattached electrons, and radio waves traveling through it, 
especially low wavelength ones, are slowed from the speed of light by 
tiny amounts. 

The amount by which the radio waves are slowed down depends on 
the density of electrons in the plasma. In interstellar space, radio 
waves are slowed down by typically one inch per second from the 
186,000 miles per second speed of light. The time delay which this 
causes between pulses observed at different radio wavelengths is 
called the dispersion measure, and depends on the square of the electron 
density multiplied by the distance to the star. By assuming a value for 
the electron density in space (it had previously been studied by other 
means), the astronomers were able to estimate the distance of the 
pulsating radio source to be 200 light years, placing it among the stars 
rather than close to the solar system or outside the Galaxy altogether. 
Though the slowdown of radio waves by the interstellar plasma is a small 
decrease from the speed of light, the distance that the waves travel is so 
large that the difference in speed causes a measurable delay of typically 
one second. 

How big is it ? 

At this time the radio astronomers were also able to make 
an estimate of the size of the pulsating star (it was this latter phrase 



which was contracted to "pulsar"). They measured the length of the 
individual pulses at a particular radio wavelength and found that each 
pulse lasted for about 16 milliseconds, only two per cent of the period 
between the pulses. Each cycle of the pulsar was a brief flash with a 
relatively long time between flashes, and whatever was causing the flash 
had to emit the light all within a time of 16 thousandths of a second. 

The duration of the flash arose from two causes. Consider two parts 
of the flashing region: let us suppose that each part simultaneously 
makes a brief mini-flash, but let us suppose that the two parts are 
separated on our line of sight by a distance, say d. At the speed of light, c, 
the rear flash takes an extra time djc to reach Earth. Therefore when the 
flashes are observed at Earth they appear as a pair of pulses separated 
by an interval equal to the time taken by light to travel the distance 
between the two parts of the flashing region. 

Clearly, if the flashing region has many components spread over a 
distance along the line of sight, the flash observed would be smeared 
out over the time that would be taken by light to travel that distance. 
If the flashes from the component parts are not really simultaneous then 
the spread of the pulse will be even more than the light-time along the 
depth of the source. Therefore, the duration of the flash of the pulsar as 
observed on Earth shows the maximum extent of the pulsar's depth. 
The pulsar observed by Bell and Hewish must be smaller than 16 light 
milliseconds in depth: less than 3500 miles. 

What sort of object could emit rapid, energetic radio pulses, yet be 
smaller than the Earth ? This is much smaller than any ordinary star, 
but it is about the size of certain very condensed stars. For many years, 
astronomers had been aware that there are some very strange stars indeed 
in the sky. Among the strangest are the white dwarfs, the end products 
of stars in which the normal energy processes have ended, leaving them 
at the mercy of their own gravity. These stars collapse upon themselves, 
forcing their atoms into a super-dense state called degeneracy, and 
resulting in the entire mass of the star being packed into a body little 
larger than the Earth. White dwarfs are not uncommon in the sky — one 



is a companion of the bright star Sirius, for example. But theoretical 
astronomers did not rest content with white dwarfs. There should be 
even stranger stars, they said, smaller and denser than white dwarfs. 
They called them neutron stars. For 30 years, neutron stars were the 
theoretical solution to a problem which did not exist: mathematically it 
had been shown that they could exist but no trace had been found of a 
real example. As evidence grew that the newly discovered pulsar was a 
natural phenomenon, the thoughts of the Cambridge astronomers turned 
to stars of this kind. But there were still more discoveries to be made. 

More pulsars turn up 

In December, Bell discovered a second pulsar. "I was 
working in the evening analyzing charts. I saw something which looked 
remarkably like the bits of scruff we had been working with. This was 
in a bit of sky which wasn't very easy for the telescope to look at, but 
there was enough to confirm that there had been scruff. 

"That particular bit of sky was due to go through the beam at one in 
the morning. It was a very cold night and the telescope doesn't perform 
very well in cold weather. I breathed hot air on it, I kicked and swore 
at it, and I got it to work for just five minutes. It was the right five 
minutes, and at the right setting. The source gave a train of pulses but 
with a different period of about one and a quarter seconds." 

Finding a second pulsar made it even less likely that the transmissions 
were artificially produced by another civilization. "There wouldn't be 
two lots signaling us at different frequencies. So obviously we were 
dealing with some sort of very rapid star. I threw up another two some- 
time in January." "Throwing up" another two had involved searching 
back through the three miles of tape which had accumulated. Now that 
they had this further evidence that the signals were from a natural 
galactic source, the astronomers felt ready to publish the news of the 
discovery of the first pulsar. They wrote the paper describing the 
observations, and submitted it to the science journal Nature eight weeks 



after the recognition that the first radio source was pulsating. Shortly 
afterwards, on February 24, 1968, the paper on the first pulsar was 
published in Nature, authored by Hewish, Bell, and their colleagues 
J. Pilkington, P. Scott and P. Collins. 

The story of the discovery of pulsars has a pattern which recurs 
regularly in the history of science, though not as repetitively as pulsars 
themselves. Built for a completely different purpose, Hewish's radio 
telescope picked out the pulsars by chance. Bell noticed the signal from 
the first pulsar because, though it resembled both spurious interference 
and the normal twinkling of radio stars, its characteristics did not quite 
fit with either because it recurred in the same part of the sky and it 
twinkled at midnight. Although she was still a graduate student and 
therefore inexperienced in astronomy, she had a receptive mind and she 
seems to have been less ready to dismiss the " scruff" as interference than 
her more experienced colleagues (there is a persistent rumor that pulsars 
had been seen by a radio astronomer before Bell, but had been dismissed 
by him). After patiently but quickly bringing out the essential charac- 
teristics of the pulsars, the Cambridge group, armed with the confidence 
that they had discovered further examples, published their data and 
were able to mention, in their discussion of it, what is now believed to 
be the kind of star responsible. 

A row flares 

Like many stories of superb scientific discoveries, this one 
has parts that could have been scripted by C. P. Snow. The Cambridge 
group was criticized for sitting on the discovery of the first pulsar for 
six months and then concealing the discovery of the further pulsars 
which they had found. Actually there were only two months between 
the recognition of the repetitive pulses from the first pulsar and sub- 
mission of their paper to the journal Nature * 

* Fred Hoyle's novel A for Andromeda, written with John Elliot and 
published five years before Bell's discovery, contains a remarkable 



Professor Sir Bernard Lovell, Director of Britain's Jodrell Bank 
Observatory said he thought " the Cambridge people had behaved with 
exemplary scientific discipline in withholding news of their discovery 
until they were satisfied about the general nature of the objects, but that 
there was no excuse for a similar delay in withholding information about 
similar objects which they had discovered." But this incident was of 
small importance compared with a later row. 

In 1974, the Nobel Prize for physics was awarded for the first time to 
astronomers. It was given jointly to Martin Ryle, Director of the radio 
observatory at Cambridge for his work in developing new kinds of radio 
telescopes, and to Antony Hewish for his "decisive role in the discovery 
of pulsars." The Nobel Prize committees are thought to investigate the 
circumstances very thoroughly when they make their awards, and they 
seem to have been satisfied that Bell's part in the discovery did not 
merit a share in the prize. 

In the following March, however, the well-known British astronomer, 
Fred Hoyle, one-time director of the Institute of Theoretical Astronomy 
in Cambridge, criticized the way in which the Nobel Prize had been 

foreshadowing of the discovery of pulsars. A large, new radio telescope 
picks up from the direction of the constellation Andromeda a "faint 
single note, broken but always continuing" like Morse code. The radio 
astronomers deduce from its constant position in galactic coordinates 
that it is not in orbit in our solar system, nor an artificial satellite. 
Unlike the real pulsars the signal turns out to be a message, with a lot 
of "fast detailed stuff" between the dots and dashes. Like the Cam- 
bridge radio astronomers, the fictional ones build a fast recorder to 
record it. A news blackout is imposed by a high-up civil servant, to be 
broken by the most individualistic of the scientists. The press sensa- 
tionalize the signal: "spacemen scare: is this an attack?" 
Perhaps this is another reason why, when pulsars were first discovered 
at Cambridge, the radio astronomers imposed a news blackout on 
themselves, avoiding misrepresentation when they were unsure of 
precisely what they were observing. 



awarded to Hewish without including Bell. According to Hoyle, the 
crucial parts of the discovery were the recognition of the signal as some- 
thing unusual, and the observation that it was keeping sidereal time. 
After this, any astronomer would have gone through the same reasoning 
process and come to the same conclusions as the Cambridge team. 
Hoyle wrote : 

There has been a tendency to misunderstand the magnitude of Miss BelVs 
achievement, because it sounds so simple, just to search and search through a 
great mass of records. The achievement came from a willingness to contem- 
plate as a serious possibility a phenomenon that all past experience suggested 
was impossible. I have to go back in my mind to the discovery of radioactivity 
by Henri Becquerel for a comparable example of a scientific bolt from the 

In a reply Hewish wrote that Bell had been carrying out a program 
initiated and mapped out by him. He said that her work as a graduate 
student had been excellent, but that it would be unjust to later graduate 
students who continued the analysis to suggest that they would not have 
discovered the pulsar themselves, had they been in her position. Of 
course, the argument that the next person down the line would have 
made the discovery anyway if X had not, applies to most scientific work. 
Very often in great discoveries there is an element of luck: being in the 
right place at the right time, with the right predecessors. No matter who 
received the Nobel Prize, Bell actually discovered the first pulsar of the 
149 now known. 

More discoveries 

After the four Cambridge pulsars were found others were 
discovered by astronomers using radio telescopes at Green Bank, W. Va. ; 
Jodrell Bank; Arecibo, Puerto Rico; and Molonglo in Australia. In a 



discovery late in 1968 which indicated the connection between pulsars 
and supernovae, a Sydney University group using the Molonglo radio 
telescopes discovered a very short period pulsar which lay in the same 
direction as a source of radio emission called Vela X. Could the two be 
linked ? Vela X had been previously identified by Douglas Milne as the 
remnant of a supernova which occurred some 10,000 years ago. Milne 
put Vela X at a distance of about 1700 light years, and the Sydney radio 
astronomers M. I. Large, A. E. Vaughan and Bernard Mills deduced 
from its dispersion measure (p. 103) that the pulsar which they dis- 
covered was at the same distance. They inferred that the pulsar probably 
was the stellar remnant of the supernova — the small star left over after 
the supernova explosion which had ejected the outer parts of the original 
star into space and made the Vela X radio source. The pulsar had a very 
short period — only 89 thousandths of a second, and each brief flash of 
radio waves lasted only 10 thousandths of a second. 

Within a few weeks, however, an even shorter period pulsar (still 
the shortest period known) was found in the center of the Crab Nebula 
which proved the connection between pulsars and supernovae. This one 
was discovered in a deliberate search of the Crab for its pulsar by 
D. H. Staelin and E. C. Reifenstein at the National Radio Astronomy 
Observatory in Green Bank, W. Va., using a brilliant method which 
exploited one of the properties that made its detection as a pulsating 
star difficult — its high dispersion. 

As explained previously, pulsar pulses, when observed at different 
radio wavelengths, arrive at Earth at different times. Now, no radio 
telescope observes at a single radio wavelength — it always detects radio 
waves arriving within a small band of wavelengths, called the bandwidth 
of the radio receiver. (This is true of a domestic radio receiver too : 
cheaper quality radios receive a wider bandwidth than more finely 
tuned, expensive ones and may receive the programs broadcast by two 
radio stations at adjacent wavelengths on the wavelength dial. The two 
stations may interfere with each other and their individual signals are 



The wider the bandwidth in a radio astronomy receiver, the more 
radio energy it receives from the sky and as a consequence it can detect 
fainter radio stars. But, if a radio star is pulsating fast and there is strong 
dispersion, a pulse from the pulsar can be received at the lower wave- 
length edge of the bandwidth at the same time as previous pulses are 
being received at longer wavelengths still in the receiver's bandwidth. 
As a result the pulses are blurred together and not distinguished or 
recognized as pulses by the radio telescope. 

Staelin and Reifenstein realized that this would occur if they were to 
search for a pulsar in the Crab Nebula since dispersion is caused by 
electrons in space and the Crab contains many free electrons. This is 
evident from its appearance, with red H-alpha filaments and overall 
synchrotron glow. They decided therefore to use finely tuned narrow 
bandwidth radio receivers so that they could observe individual pulses, 
compensating for each receiver's individual lack of sensitivity by coupling 
them all together in a bank of 50, and looking for a particular pulse as it 
was received in turn by each of the receivers. It turned out, when they 
in fact discovered the Crab Nebula pulsar, that it took i\ seconds for a 
pulse from the pulsar to sweep through all 50 of their receivers, because 
of the dispersion. The period of the pulsar turned out to be just 33 
thousandths of a second. 

The two shortest period pulsars then known had been discovered to 
be associated with supernova remnants. To this day no other pulsar of 
longer period has been unambiguously associated with a supernova 
remnant. Searching for a reason, astronomers hypothesized that the 
youngest pulsars had the shortest periods and pulsars slow down as they 
age. Only young and hence short period pulsars would be found 
associated with supernova remnants, as the remnants associated with old 
pulsars whould have dissipated into space, fading from view. 

Almost instant confirmation of this idea came from the discovery that 
the shortest period pulsar, the Crab pulsar, was not completely regular. 
Within a month the period had been measured accurately enough to 
demonstrate that it had increased by one millionth of a second. This 


T T 

Observations of the light flashes from the Crab pulsar made with the 4-meter 
Anglo-Australian Telescope. The main flashes {marked on the top scale) are 
0.033 seconds apart. Between the main flashes can be seen the so-called 
inter pulse, smaller flashes from a weaker beam on the rotating neutron star, 
following its main beam by 0.013 seconds. Doubled flashes are relatively common 
among the faster pulsars. 

steady increase meant that the Crab Nebula pulsar was slowing down 
at a rate which indicated a lifetime of just about 1000 years — near 
enough the age of the Crab Nebula as determined from the Chinese 
observations of the supernova of 1054! By the time this was known late 
in 1968, the four pulsars discovered by Bell and her Cambridge col- 
leagues had been under observation both at Cambridge and Jodrell Bank 
for more than a year, and their periods were becoming known with 
greater and greater precision. All four were found to be slowing down, 
but some 10,000 times more gradually than the Crab pulsar. At the 
present time 83 of the 149 pulsars now discovered are known to be 
slowing down — not one is known to be speeding up — further indication 
that the rotation of pulsars slows as they age. 

Seeing the Crab pulsar flash 

Although radio observations immediately showed clearly 
that the pulsar discovered by Staelin and Reifenstein was in the Crab 
Nebula, its position could not be measured well enough with radio 
telescopes to determine precisely which star was the pulsar. Indeed, 
although many attempts had been made to identify the first radio pulsars 
with particular stars — and there were several false alarms — no radio 
pulsar had been found to be identical to any optically visible star. These 



earlier disappointments may explain the casualness of the efforts made 
by optical astronomers with access to large telescopes to identify the 
Crab Nebula pulsar. One astronomer used the 98-inch Isaac Newton 
telescope in Britain, six days after the periodicity of the radio Crab 
pulsar was announced, to observe light from the central regions of the 
Crab Nebula, but left his data, which was in the form of punched tape 
readable only in a large computer, unanalyzed for months with no ink- 
ling of the discovery hidden within. 

Then a team of three astronomers and physicists at the Steward 
Observatory in Tucson Arizona used the much smaller 36-inch telescope 
there in January 1969 to look in the center of the Crab Nebula for light 
flashes with the radio period. In their experiment W. J. Cocke, M. J. 
Disney and D. J. Taylor used a small computer to add thousandths of 
the flashes together, and — this turned out to be a significant difference — 
they had the results of the computer summation available to them in 
graphical form then and there, while the experiment was in progress. 
As the pulses came in from the Crab pulsar, they were displayed on a 
multi-channel recorder — a glowing screen on which pulses appeared as 
a hump on a line, growing as the astronomers watched. They were able 
to sweep the telescope over the central regions of the Crab Nebula to see 
where the pulses came from and were able to say that they came from 
the vicinity of the two central stars, the northeast and southwest stars 
named by Baade. They could not say which, for sure, but guessed the 
southwest star as this had been the one backed by Baade and Minkowski. 

The Steward Observatory experimenters' guess was confirmed in a 
clever experiment by two astronomers from Lick Observatory, Joe 
Miller and Joe Wampler. They used a TV camera attached to the 
120-inch Lick telescope, peering at the pulsar through a rotating shutter 
which was made so that the open periods were separated by a time very 
nearly the same as the period of the pulsar. Similar devices, called 
stroboscopes, are used to view rapidly rotating machinery; the strobo- 
scopic effect is also familiar as a slowing of the motion of a wagon wheel 
when filmed by a movie camera which exposes film frame by frame, 


A pulsar flashes. This remarkable pair of photographs of the Crab Nebula pulsar 
was taken with the Lick 120-inch telescope using a television system looking 
through a stroboscopic shutter. The top photo was exposed when the pulsar was 
at its brightest during its flash, the lower when it was almost invisible. The 
other stars nearby have a constant brightness. This is a Lick Observatory 
photograph by E. J. W ampler and J. Miller. 



through a shutter. When the pulsar was flashing at the same times that 
the shutter was closed, the Lick astronomers saw nothing. But when it 
flashed at the times the shutter was open they immediately saw that the 
pulsar was Baade's southwest star, right at the center of expansion of the 
Crab, and at the center of activity of the light ripples seen by him. 

Astronomers had to look no further for the powerhouse which drives 
the Crab's expansion and generates the electrons to produce its syn- 
chrotron radiation. 

A really exciting moment 

If a pulsar did not exist in the Crab Nebula it would be 
necessary to invent one. Indeed it was necessary, for the year before the 
discovery of pulsars, F. Pacini suggested that a rotating neutron star 
was the present-day source of the fast electrons in the Crab Nebula. 
The discovery of the Crab Nebula pulsar showed that this was possible, 
and the further discovery that the pulsar was slowing down proved it, 
as the star's spin-energy lost in the slowdown was closely equal to the 
total amount of light and radio energy radiated by the synchrotron 
process. The whole flow of energy within the nebula became clearer. 

Thomas Gold of Cornell University recalls the moment when he too 
realized that the Crab Nebula was powered by a rotating neutron star : 
"Let me just recount for you our excitement at Cornell when we had 
just obtained, at our observatory at Arecibo, a slowdown rate for the 
Crab for which we had been looking. We had expected a supernova to 
give rise to a neutron star, which if found, was expected to be slowing 
down. Therefore, when the pulsar in the Crab was discovered, we 
immediately started looking for a slowing down of the pulsations since 
such a short period pulsar should slow down fast. When we got the 
information, we immediately worked out the rate of change of energy 
of a rotating neutron star, having been previously very impressed with 
the very high total energy content in the rotation of the object. 

" I remember doing the completely simple calculation of what the 



rate of change meant in terms of the energy output, and meanwhile 
sending an assistant to the library because I no longer remembered the 
quoted figures for the total energy output of the Crab Nebula that had 
been calculated by Shklovsky years before. When he returned in a few 
minutes with various references giving estimates of the luminosity of 
the Crab we found that the figure on my pad and the other in the book 
were the same, namely io 53 ergs per second. That was a really exciting 
moment. I realize that we can't be quite sure that this is the right 
number, but still, it isn't often that you hit it off like that with a com- 
pletely theoretical calculation of such a far-fetched thing as the structure 
of a neutron star." 

Superstar . . . 

There was however still an unexplained problem. The pul- 
sar itself was too faint. Responsible for the vast amount of energy pouring 
into the Crab and manifest as accelerating filaments and as synchrotron 
radiation, the pulsar itself emitted just ten millionths as much energy as 
visible radiation. It was scarcely credible that the pulsar would pulse 
on its own behalf such a small amount of energy, while passively passing 
on such huge amounts, any more than it is credible that a showbusiness 
superstar would accept a small paypacket while generating millions of 
dollars worth of business. Searching for the Crab pulsar's piece of the 
action, NRL and Columbia University X-ray astronomers launched two 
rockets in March 1969 and found X-ray pulsations. This prompted 
other astronomers to search through and re-analyze old data which they 
had not previously thought to examine for X-ray pulsations. These 
pulsations occur exactly in step with the optical flashes, and the period 
of both is the same. 

The energy which the Crab pulsar puts into X-radiation is 20 times 
the energy it emits optically and 20,000 times the energy it pulses as 
radio radiation. 


. . . and co-star 

Until 1975 the Crab pulsar was the only pulsar detected as 
an optical or X-ray pulsar as well as a radio pulsar. The search for other 
examples had proved fruitless. The most likely pulsar to be discovered 
as an optical or X-ray pulsar was the Vela pulsar in the Vela supernova 
remnant. Before 1975 it was the second fastest pulsar known and so had 
more rotational energy to convert to light and X-rays than any other 
pulsar apart from the Crab. X-rays have in fact been detected from a 
star near the Vela pulsar, and there was one report in 1973 that the 
X-rays pulsed, but more sensitive X-ray telescopes failed to see any 
pulsations. In 1975 however, the Small Astronomy Satellite, SAS-2, 
detected pulses from the Vela pulsar at very energetic X-ray energies 
(gamma rays in fact). 

Optical astronomers redoubled their efforts to find visible light from 
the Vela pulsar. Though four times nearer than the Crab and in a less 
dusty region of the Galaxy, the Vela pulsar was expected to be much 
fainter than the Crab because ten times older and three times slower. 
The optical search was hampered by the disagreement among radio 
astronomers about the precise direction of the pulsar. Within the zone 
of uncertainty there were too many faint stars to be examined in the 
detail necessary to pick it up. However after radio astronomers Dick 
Manchester, Miller Goss and Bruce McAdam had pinpointed the 
position of the Vela radio pulsar with an accurate radio telescope at 
Fleurs, near Sydney, Australia, the radio astronomers joined forces with 
radio-astronomer-turned-optical-astronomer F. Graham Smith, and 
his team from the Royal Greenwich Observatory, together with four 
optical astronomers at the Anglo-Australian Observatory, Pat Wallace, 
Bruce Peterson, John Danziger and Paul Murdin. In 1977, using the 
newest large optical telescope in the southern hemisphere, the 154-inch 
Anglo-Australian Telescope, this large team found the Vela pulsar after 
ten hours integration on the right place in the sky. Light flashes from 
the Vela radio position were stored in a computer, integrated spon- 



taneously in an analysis at the telescope, and remembered for subsequent 
more refined analysis afterwards. The Vela pulsar, 10 billion times fainter 
than bright naked-eye stars, is one of the faintest stars to be seen by 
optical astronomers. 

The Vela pulsar differs from the Crab in that, after accounting for the 
slowdown of the radio pulse in space, its double gamma ray and optical 
flashes trail behind the single radio pulse, whereas the Crab emits its 
radio, X-ray and optical pulses simultaneously. It is too soon to say how 
this helps us to understand the way pulsars shine. 


VI I The search for supernova 

remnants: Looking for other 

¥ Then 

'hen a star goes supernova, it leaves traces. It 
would be wrong to call these remains the dead body of the star because, 
in the case of the Crab at least, the remains are still active, emitting as 
much energy as a luminous star. Are there other supernova remnants ? 
What about Tycho's and Kepler's stars, and all the other historical 
supernovae ? 

After Walter Baade had studied the light curve of Kepler's star, he 
realized that it must have been a supernova and in 1947 he searched for 
its remnant. Using what was then the world's largest telescope, the 
100-inch at Mount Wilson, he photographed the suspect area using a red 
filter to isolate the red light expected from the hydrogen in the nebula. 
At the position of Kepler's supernova he found a few wispy filaments of 
gas, the last traces of the exploding star of 1604. 

Baade applied the same principles to a search for the remnant of 
Tycho's supernova of 1 572. He was unsuccessful— photographs centered 
on Tycho's position for the supernova showed stars but not nebula. 

Then in 1952 Robert Hanbury Brown and Cyril Hazard searched for 
Tycho's supernova remnant with a 218-foot radio telescope at Jodrell 
Bank (not the fully steerable 250-foot telescope but a simpler telescope 



fixed to the ground). They found a powerful radio source near the right 
place, its position being tied down by a Cambridge group using an 
interferometer. It turned out that the position Tycho gave had the 
unexpectedly large error of 4 arc minutes — over a tenth of the diameter 
of the Moon. Though this had caused Baade to look somewhat off the 
correct place for the nebula, the real position was still on the edge of his 
plates and he would probably have seen the nebula if it were bright 

However, the remnant of Tycho's supernova is very faint indeed and 
photographs by R. Minkowski with the 200-inch Mount Palomar tele- 
scope show just a few wisps of nebula near the radio source. The radio 
source has been mapped by the one-mile Cambridge radio telescope. It 
turns out to be beautifully circular in shape with a ring of bright radio 
emission, the edge of a bubble of material thrown violently into the 
interstellar gas surrounding the supernova, colliding with it, and heating 
the gas to a billion degrees. This is why so little light is seen — the gas 
is too hot for hydrogen atoms to form and, in doing so, emit light. 

No astronomer doubts that this is the remnant of Tycho's supernova. 
Why then is it so far from the position Tycho measured for the star ? 
The measurements that he made on the nova with his new sextant 
extended over many months, but there is no record of Tycho using it for 
any of his subsequent observations, known to be very accurate. According 
to astronomer David H. Clark and historian F. Richard Stephenson the 
discrepancy can be explained purely as a small calibration error in the 
angular scale on the sextant — as if Tycho were measuring meters with a 

yard stick. 

Once the remains of the supernovae of 1054 (the Crab Nebula), 1572 
(Tycho's supernova) and 1604 (Kepler's supernova), had been identified 
with the wispy supernova remnants, astronomers began to look for other 
examples. They were hoping to find the remnants of supernovae which 
had not been recorded, either because they were too faint to be seen or 
because they occurred in prehistoric times so that no records of the 
supernovae have come down to us. 



^^Ur-> .? 

Though the visible remains ofTycho's supernova are faint, this gas shell still 
glows at radio wavelengths. This is a radio contour map, in which the high 
number contours refer to peaks of radio brightness, which is greatest around the 
edge. Hatching shows optically-visible filaments. 

The traces of ancient supernovae 

In all three cases of the well-established historical super- 
novae, radio sources were observed at the positions of the supernovae. 
These radio sources could be distinguished from other radio sources 



that have been recorded, firstly because supernova remnants have a very 
characteristic shape and are as obvious to radio astronomers as planetary 
(ring) nebulae are to their optical colleagues. In fact, the basic causes 
are the same. When a low-mass star has to shed excess mass, it puffs it 
off in a shell, a spherical bubble, which is then visible as a ring-shaped 
planetary nebula glowing in the optical region. A supernova explosion, 
though much more dramatic, throws off a similar, easily identifiable, 
radio shell. Second, the radio spectra of supernova remnants are 
different from those of other radio sources. Radio source spectra are less 
detailed than the spectra of optical sources. Radio astronomers measure 
the source's intensity at a number of frequencies, plot a graph of the 
results, and look at the shape of the curve produced. This varies with 
the way the radiation is being created. Broadly speaking, there are two 
types of mechanism for producing radio waves, thermal and non-thermal. 
Thermal sources are ones in which hot gas is involved : the radiation is 
produced in the same way as the optical light of stars, by encounters 
between electrons. In thermal sources relatively slow electrons in a 
plasma are encountering and mutually repelling one another, changing 
their motion at random. Whenever an electron changes its motion it 
emits radiation, so the plasma radiates radio waves with a balanced 
proportion of long and short wavelengths which is characteristic of 
thermal sources. 

Supernova remnants, however, are examples of non-thermal sources. 
Relatively fast-moving electrons are colliding not with each other but 
with the magnetic field embedded in the plasma. In fact the electrons 
spiral along the magnetic field, emitting radio radiation as they go — the 
synchrotron process. The spectrum of radio radiation which they emit 
depends on the fine details of the strength of the magnetic field and the 
like; therefore although the balance of long and short radio wavelengths 
is not unique it is usually sufficiently different from the thermal radio 
sources to be distinguishable. 

The latest catalogs of supernova remnants, that is, of sources which 
are extended and have non-thermal radio spectra, list approximately 



ioo supernova remnants. The remains of just over two dozen are bright 
enough to have been detected by optical astronomers as well as radio 

The most intense radio source in the sky is in fact a supernova 
remnant. In the constellation Cassiopeia, it is known as Cassiopeia A, 
Cas A for short. It was even visible but unrecognized on Reber's first 
maps of the sky in radio waves, and it was rediscovered in 1948 by 
Martin Ryle and F. Graham Smith in Cambridge. Smith determined 
its position accurately and, in 1954, Baade and Minkowski discovered the 
optical counterpart of Cas A. They found an almost complete shell of 
knots and filaments approximately four arc minutes across. 

Some of these knots and filaments were shown by Baade and Minkow- 
ski to be expanding from a central position although no star was visible 
there, so that there is no neutron star visible. The knots appear to come 
and go, with only one third of the knots being visible for more than 12 
years. They are probably produced by parts of the supernova plowing 
through stationary interstellar cloud banks. As they strike the cloud 
banks, atoms in the knots become what physicists term "excited." 
Electrons in atoms of oxygen, sulfur and argon are moved to high energy 
states. When they drop back to their original low energy states, they 
emit spectral lines of these elements. 

The ejection velocity and momentum in the knots is so high that 
collisions do not appear to have slowed the knots significantly since they 
were first ejected. Some very fast-moving knots are speeding away 
from the central part of Cassiopeia A at a speed of 9500 km/s. From the 
expansion velocity, Sidney van den Bergh and William W. Dodd have 
deduced that the supernova which produced Cassiopeia A took place 
within eight years of 1667. 

The strange thing is that no supernova is recorded in Chinese, 

* The twenty-seventh was found by Andrew Longmore, David Clark 
and Paul Murdin during the preparation of this chapter. In the 
constellation of the Southern Cross, it has the unexciting name of 
G296. 1-0.8. 



table II Some supernova remnants 








Cassiopeia A 

i66 7 ± 


R X S 

Kepler's sn 



R O H S 

Tycho's sn, 3C 10 

!57 2 


R O X H S 




R H D 

Crab, Taurus A, 3c 144, mi, 

ngc 1952 



R X H P D 

p 1459-41 



R X H S 

MSH 14-63 




R H S 

Puppis A 



R X S 

Vela x 




Cygnus Loop, Veil Nebula 



R O X S 

IC 443 



R X S 

Key to description : 

R = radio remnant 

H = historical supernova 

= optical remnant 

s = shell 

P = pulsar known 

D = disk 

x = X-ray emitter 

Korean or Japanese chronicles at this date, which suggests that when this 
Cassiopeia supernova occurred, it was fairly faint. Its distance is com- 
parable with that of Tycho's supernova, namely, just over 9000 light 
years. It may have appeared fainter than Tycho's supernova because it 
occurred in a very dusty region of the Galaxy and its light was obscured. 
Very detailed radio maps of Cassiopeia A have been made by Ivan 
Rosenberg, using the one-mile radio telescope at Cambridge. These 
maps show that Cassiopeia A is a very well defined shell supernova 
remnant although the shell has broken up into lumps in places. The 



shell appears donut shaped when seen projected on the sky, because 
where the line of sight passes along the edges of the shell, we are looking 
through a greater thickness of gas than at the center. 

The force of the collision of pieces of the exploding supernova with 
the interstellar medium pervading the area nearby heats the gas very 
strongly, to many millions of degrees. At such temperatures gas emits 
X-rays and, indeed, these have been detected coming from many 
supernova remnants (besides the Crab Nebula which, as we saw in 
Chapter V, emits X-rays by the synchrotron process). The remnants of 
Cas A and Tycho's and Kepler's supernovae are all in this class. Most 

An X-ray '■''photograph'''' of the Cygnus Loop supernova remnant, prepared 
from data obtained in the rocket flight of an X-ray telescope operated by Saul A. 
Rappaport and his MIT colleagues. The brighter areas {paler shades of gray) 
represent the most intense regions of X-rays. The remnant has a distinct shell 
structure and a central source, possibly associated with an unseen neutron star. 


supernova remnants are too small in extent to be perceived as anything 
but a spot of radiation by the current generation of X-ray telescopes, 
whose ability to see detail is poor. The largest supernova remnants such 
as the Cygnus Loop and Vela x, however, can clearly be seen as shells 
of X-ray emitting gas, roughly similar to the optical and radio pictures. 
Cas A is probably shell-like at X-ray wavelengths too. 

The Gum Nebula and the Vela pulsar 

The largest nebula in the sky is called the Gum Nebula. It 
was discovered by a young Australian astronomer, Colin Gum, whose 
career was cut short by a fatal skiing accident at the age of 36. For his 
Ph.D. thesis, Gum had photographed the whole of the Milky Way as 
it is visible in the southern hemisphere. He used a red filter so that the 
photographic plate would pick up predominantly the deep red spectral 
line emitted by hydrogen, H-alpha. Because of the predominance of 
hydrogen in the interstellar medium, H-alpha is usually the strongest 
spectral line emitted by nebulae. At the same time, the light of other 
objects in the heavens, including the glow of our own atmosphere, is 
fairly weak at the red end of the spectrum. By preventing all but red 
light from reaching the photograph, it is possible to isolate H-alpha 
emitting objects from the others, and give longer exposures without 
white light from other objects flooding the photograph. Gum made a 
mosaic of several of his photographs of the constellations Vela and 
Puppis, and the nebula which now bears his name at once became 
apparent. It has a diameter of about 30 degrees on the sky. This means 
that if a person's eyes were sensitive enough to perceive the faint H-alpha 
emissions from the Gum Nebula, it would fill half the area of the sky 
that one eye can see at a time. 

Within the Gum Nebula lie the two stars Zeta Puppis and Gamma 
Velorum. Gum thought that these two stars were emitting enough 
ultraviolet light to be able to separate, or ionize, any hydrogen atoms in 
the vicinity into electrons and protons. The result would be that when 


,, s i Mi?" • 

Photograph of the Gum Nebula by John Brandt, Robert Rosen, J. Thompson 
and D. J. Ludden, made with red light in a wide-angle camera to show H-alpha 
emission over an area 40 degrees square. The Vela supernova remnant and Vela 
pulsar are below left of center. Absorption bands of nearby dust (dark zones) 
obscure the millions of stars of the Milky Way as it runs left to right across the 
center of the picture, from Carina through Vela, Pyxis and Puppis. The 
appearance of the Gum Nebula in this picture is consistent with it being caused by 
the flood of photons released in a supernova explosion. 

these electrons and protons recombined into hydrogen atoms, they 
would emit the H-alpha light that he saw coming from the nebula. The 
recombined atoms would almost immediately be re-ionized by the two 
stars' ultraviolet light, to recombine again in a repetitive cycle; the 
nebula is constantly regenerated. The two stars are 1500 light years 
away from the Earth, while the Gum Nebula is approximately 1000 light 
years in radius, so that the Sun lies close to the nearer edge, not quite 
within the Gum Nebula, but not far away. 



In 197 1, a group of four astronomers, John Brandt, Theodore 
Stcchcr, Steve Maran and David Crawford, calculated that Gamma 
Velorum and Zeta Puppis did not produce enough energy to be able to 
split apart the large number of hydrogen atoms within such a large 
volume of space. They said instead that the hydrogen atoms of the 
Gum Nebula had originally been split apart, not by the steady radiation 
from the stars within it, but by a burst of radiation from the explosion 
of a supernova some 10,000 years ago. The ionized gas will ultimately all 

Many thousands of years after a supernova explosion all that remains is a 
spherical shell. This is the Veil Nebula in Cygnus, one of the brightest supernova 
remnants. The wisps glow visibly at the edge of a radio shell. Radio astronomers 
have found no trace of a pulsar at the center of this object, which is about 
20,000 years old. Hale Observatories photograph. 

The brighter half of the filamentary shell of the Vela supernova remnant fills 
most of this picture taken in the red light ofll-alpha by John Meaburn and Ken 
Elliott using the 48-inch UK Schmidt Telescope at Coonabarabran, Australia. 
The Vela pulsar lies at center right, though far too faint to show. The milky 
diffuse patches, most prominent at lower left, are parts of the Gum Nebula, 
recombining hydrogen atoms which were possibly separated by the initial flash of 
radiation from the Vela supernova some 10,000 years ago. 

The motley collection of small fragments in the upper left is the visible 
remains of another supernova remnant, the strong radio source Puppis A. Science 
Research Council copyright. 





recombine to make hydrogen atoms and the nebula will cease to shine, 
but the recombination process takes many hundreds of thousands of 
years because the distance between the electrons and protons is so vast 
that a wandering electron seldom passes close enough to a proton to be 
grabbed and made into a hydrogen atom. 

According to this interpretation, the Gum Nebula is not a nebula in 
the usual sense since the stars present are insufficient to keep the nebula 
constantly ionized. It is a new kind of nebula created suddenly by a 
supernova explosion and gradually returning to its normal state, not 
being renewed. The Gum Nebula is a fossil left behind by an event 
long past. 

One must admit that many astronomers are unhappy with the 
interpretation of the Gum Nebula as the fossil of a relatively recent 
supernova explosion. Some argue that the Gum Nebula may be a 
remnant from a supernova that exploded a million years or more ago. 
They argue that its size has been over-estimated and that its two stars, 
Gamma Velorum and Zeta Puppis are indeed capable of ionizing the 
hydrogen within it in order for it to emit H-alpha light. 

Is there evidence that a supernova did occur somewhere near the 
center of the Gum Nebula about 10,000 years ago? The answer is yes. 
In the center of the Gum Nebula is the smaller nebula known as the 
Vela x supernova remnant (snr). This snr was discovered by Douglas 
Milne in 1968 using a radio telescope, and we first came across it in 
Chapter VI, in connection with the discovery of pulsars. Photographs 
of the area show a filamentary nebula looking just like other well-known 
snrs, such as the much-photographed Veil Nebula in the northern sky. 
A little off-center from the Vela x snr lies the Vela pulsar. The Vela 
pulsar's period is just less than to second, and is changing at such a rate 
that it doubles its period every 11,000 years. This can be taken as a 
measure of the age of the pulsar, which is consistent with the charac- 
teristics of the Vela x snr and with estimates of its age from other 


VI 1 1 Types of supernovae : 
gathering the evidence 


.stronomers are convinced of the connection between 
supernovae, pulsars, and supernova remnants. But of course a supernova 
does not appear from nothing. What stars become supernovae ? It turns 
out that there are at least two kinds of stars which produce supernovae: 
only one kind has been confidently identified. One answer emerged when 
the supernovae themselves were studied and it was Fritz Zwicky who 
organized the effort required to determine it. When he first set up the 
Palomar Supernova Search, Zwicky arranged a cooperative effort to 
follow the discovery of a supernova by detailed study. Walter Baade 
would measure the light curves and Rudolph Minkowski and Milton 
Humason would obtain spectra of the supernovae using the largest 
telescopes then available, the 6o-inch and ioo-inch Mount Wilson 
telescopes. Like other professionally-used astronomical telescopes, these 
are scheduled in advance, so that a particular night is pre-assigned to a 
specific astronomer to work on a project whose value he has justified to a 
committee, in competition with other observers who also wish to have 
the use of the telescope. Because present-day large telescopes are over- 
subscribed by a factor of two or three, the competition is fierce. There 
is not enough time on the telescope to satisfy all astronomers. 

By their character supernovae are unpredictable, therefore one cannot 
pre-assign time on telescopes for their study. Zwicky persuaded the 



Mount Wilson Observatory Director, George Hale, to set up an over- 
ride on observing time on the big telescopes at Mount Wilson such that 
the scheduled astronomer had to yield time for Baade or Minkowski to 
observe supernovae bright enough to be worth studying. (An override 
system for bright supernovae still operates at the 98-inch Isaac Newton 
Telescope in the U.K. An informal system now operates at Palomar and 
the scheduled astronomer is not obliged to give way, except insofar as he 
wishes to continue to have good relations with his colleagues.) 

Two kinds of supernovae 

From studies of spectra of the first supernovae discovered 
in Zwicky's supernova search, Rudolph Minkowski found in 1940 that 
most supernovae discovered (three quarters) are of a single type, which 
he called Type I. The remainder were different from Type I, though not 
all alike, and they are known as Type II. (Zwicky has been able to 
recognize further Types III, IV and V, but these types consist of iso- 
lated examples and not all astronomers are convinced that they are 
distinct kinds of supernovae.) 

The Type I supernova has an instantly recognizable light curve, 
consisting first of the sudden spectacular brightening, a quick decline in 
less than a month, and then a slower fade-off. In the case of the super- 
nova of 1937 in the galaxy ic 4182, the brightest supernova of this 
century, astronomers followed this fade-off for more than 600 days 
during which time the supernova faded without showing signs of 
stopping its decline. This supernova light curve is considered the 
prototype of Type I. 

Wondering whether the supernovae seen in our Galaxy were of Type I 
or not, Walter Baade plotted in 1943 the light curves of Tycho's and 
Kepler's supernovae of 1572 and 1604 on the same graph. He was helped 
by the almost unbelievably accurate observations of the stars' bright- 
nesses made by Brahe, Kepler and other 17th century astronomers, with 
no instruments to help them apart from their eyes. 






£ & 










(sapraiuSBui) ssaujqSijg 



1 « 
O 45 


Light curve of Brahe's supernova 

Astronomers describe the brightness of a star by a number 
they call its magnitude, but rather confusingly the magnitude scale 
seems to run the wrong way. Bright stars are said to be of the first 
magnitude, and the faintest ones which can be seen with the naked eye 
are sixth magnitude. Venus and Jupiter are brighter than first magnitude 
stars, and so actually have negative magnitudes. With a telescope one 
can perceive fainter and fainter stars which have bigger and bigger 
magnitudes. The faintest star whose light has been detected is the Vela 
pulsar at magnitude 25. The brightest is the supernova of 1006 which, 
at magnitude — 8 to — 10, was 100 million million times as bright as the 
Vela pulsar. 

Brahe's observations of the supernova of 1572 are drawn up in a form 
similar to that used by present-day astronomers to measure the bright- 
ness of variable stars. They cause starlight to fall into a photometer, an 
instrument which produces an electric current indicating the star's 
brightness. By pointing the photometer at constant stars whose bright- 
ness or magnitude has already been determined and then at the variable 
star they see how the constant stars match with the variable star. This 
is just what Tycho Brahe did, except that he used his eye, not an 
instrument, to compare the brightnesses. A feature of this method is that 
if at a later date more accurate magnitudes of the constant stars become 
available, the observations of the variable star can be reconverted into 
equally accurate magnitudes. 

Thus, when Tycho says that in July and August 1573 the nova was 
equal to the principal stars in Cassiopeia, he himself deduces that it was 
third magnitude, this being the standard of brightness set down on the 
basis of those stars' appearance to Ptolemy and published in his book 
Almagest (ad 144). The four stars in Cassiopeia which Ptolemy says are 
third magnitude are Alpha Cassiopeiae, whose magnitude measured 
with a photometer is 2.47, Beta Cassiopeiae at 2.42, Gamma Cassiopeiae 
at 2.25, and Delta Cassiopeiae at 2.80. The average is close to magnitude 
2.5, so this was the magnitude of the supernova between July and 


TYPES OF supernovae: gathering the evidence 

August 1573 in the modern system of measuring magnitudes. The fact 
that the observations have been converted into modern stellar magni- 
tudes 400 years -after they were made, rather than the day after, is 
immaterial and a surprisingly accurate light curve of Brahe's supernova 
can be drawn up. He wrote: 

When first seen the nova outshone all fixed stars, Vega and Sirius included. 
It was even a little brighter than Jupiter, then rising at sunset, so that it 
equaled Venus when this planet shines in its maximum brightness. The nova 
stayed at nearly this same brightness through almost the whole of November. 
On clear days it was seen by many observers in full daylight, even at 
noontime, a distinction otherwise reserved to Venus only. At night it often 
shone through clouds which blotted out all other stars. 

However, the nova did not retain this extraordinary brightness throughout 
its whole apparition but faded gradually until it finally disappeared. The 
successive steps were as follows: 

As already stated, the nova was as bright as Venus in November [1572]. 
In December it was about equal to Jupiter. In January [1573] it was a little 
fainter than Jupiter and surpassed considerably the brighter stars of the 
first class. In February and March it was as bright as the last-named group 
of stars. In April and May it was equal to the stars of the second magnitude. 
After a further decrease in June, it reached the third magnitude in July 
and August, when it was closely equal to the brighter stars of Cassiopeia, 
which are assigned to this magnitude. Continuing its decrease in September, 
it became equal to the stars of the fourth magnitude in October and Novem- 
ber. During the month of November, in particular, it was so similar in 
brightness to the nearby eleventh star of Cassiopeia that it was difficult to 
decide which of the two was the brighter. At the end of 1573 and in January 
1574 the nova hardly exceeded the stars of the fifth magnitude. In February 
it reached the stars of the sixth and faintest class. Finally, in March, it 
became so faint that it could not be seen any more. 

Walter Baade condensed this description to modern magnitudes and 
produced Table III. He then turned to Kepler's supernova of 1604. 



T A B L E 1 1 1 Magnitudes of Ty chefs supernova 

Days after 









Almost as bright as Venus 




About as bright as Jupiter 





A little fainter than Jupiter 




Equal to brighter stars of first mag 

+ 0.3 



Equal to second magnitude stars 

+ 1.6 


2 59 

Equal to Alpha, Beta, Gamma, 
Delta Cassiopeiae 

+ 2.5 



Equal to stars of fourth magnitude 

+ 4.0 



Equal to Kappa Cassiopeiae 

+ 4.2 




Hardly brighter than fifth mag 

+ 4-7 




Equal to stars of sixth magnitude 

+ 5-3 



Became invisible 

+ 6.0 

Light curve of Kepler's supernova 

The light curve of Kepler's supernova can also be well 
determined from the 17th century observations because Mars, Jupiter 
and, later, Saturn, provided useful comparisons. The maximum bright- 
ness was somewhat brighter than Jupiter in mid October 1604, and it 
was still almost as bright as Jupiter when it disappeared near the Sun in 
November. In January, when it reappeared, it was only about as bright 
as Antares. It continued to become fainter until it reached the fifth 
magnitude in October 1605. By the following spring, it was no longer 


TYPES OF supernovae: gathering the evidence 

visible to the naked eye. Throughout its appearance Kepler made a 
series of comparisons between it and other stars, possibly modeled on 
Brahe's observations of the supernova of 1572. 

table iv Magnitudes of Kepler's supernova 

Days after 









Oct 8 


Not seen 


+ 3 
or more 

Oct 9 


As bright as Mars 

The physi- 


Oct 10 


Like Mars 



Oct 11 


Still brighter than Oct. 10 

The physi- 


Oct 12 


Almost as bright as Jupiter 



Oct 15 

— 2 

As bright as Jupiter 

The physi- 
cian and 

— 2.2 

Oct 17 

Much brighter than Jupiter 




Jan 3 


Brighter than Alpha Scorpii 



Jan 13 


Brighter than Alpha Bootis 
and Saturn 



Jan 14 


About as bright as Mars in 
Oct 1604 



Jan 21 


As bright as Alpha Scorpii 





As bright as Alpha Virginis 



Mar 20 


Not much brighter than 



Zeta and Eta Ophiuchi 



Days after 


brightness Description 



Mar 27 





Mar 28 


Not much brighter than 
Eta Ophiuchi 



Apr 12 


As bright as Eta Ophiuchi 



Apr 21 





Aug 13 


As bright as Xi Ophiuchi 



Aug 29 


About as bright as Xi 



Sep 13 


Fainter than Xi Ophiuchi 



Oct 8 


Fainter than Xi Ophiuchi, 
difficult to see 



Drawing the light curves of Tycho's and Kepler's supernovae on the 
same graph as the light curve of the supernova he observed in I c 4182, 
Baade found that all three objects are the same kind of supernova- 
Type I. 

The Crab— Type I or II? 

About the supernova of 1054, the Crab Nebula supernova, 
there is, however, some doubt. The data are scanty but it seems that 
after sn 1054's fairly quick decline to magnitude -3.5 it faded to 
magnitude +5 in 630 days, a drop of 8.5 magnitudes, compared with 
the norm of 10.5 magnitudes for the prototype Type I. Thus, writes 
Minkowski, " the difference of 2 mag between the supernova of 1054 and 
supernovae of Type I is not conclusive evidence in view of the many 
uncertainties, but it tends to contradict the interpretation of the super- 
nova of 1054 as Type I and certainly does not make it mandatory". 
This is important in trying to explain why the remnants of Tycho's and 


TYPES of supernovae: gathering the evidence 

Kepler's supernovae are so different in their appearance from the Crab 
Nebula: it would be easier to understand if the former were Type I and 
the latter Type H. 

The true brightness of a supernova 

The brightness of all Type I supernovae at their maximum 
is probably the same, insofar as this can be tested. There are three 
uncertainties which have to be considered. The first is that many super- 
novae are on the decline when first discovered and there is no way of 
determining what their maximum brightness is. Only a supernova whose 
maximum brightness "is well established can be used in this test. 

Second, supernovae are at varying distances from Earth so that their 
brightness as measured depends on how far away they are. Astronomers 
have to correct their measurement of the brightness of the supernova by 
calculating how much brighter it would appear if placed at some standard 
distance from Earth. (For various historical reasons this standard 
distance has been chosen as ten parsecs, about 32.5 light years.) To make 
the calculation they have to know at what distance the supernova lies. 
This can be done for the nearest galaxies by comparing the brightness of 
certain kinds of variable stars in those galaxies (cepheids) with cepheids 
in our Galaxy. Often, however, no individual stars can be distinguished 
in a galaxy in which a supernova occurs (save the supernova itself) and 
astronomers fall back on the so-called Hubble redshift relation. 

Edwin Hubble found in 1929 that on average all galaxies were 
receding from our Galaxy at speeds such that the more distant galaxies 
recede faster, and their spectra are redshifted more, because of the 
Doppler effect. According to the best recent determinations 55 km/s is 
added on average to a galaxy's recessional speed for every million parsecs 
(3.25 million light years) distance it is from Earth. Thus, if a galaxy's 
speed is measured by the Doppler effect to be 5500 km/s, its distance is 
around 100 million parsecs. A supernova of Type I might have a maxi- 
mum brightness of 16 in this galaxy, but when brought to a distance of 



just io parsecs (10 million times closer) it would be 10 million squared 
times brighter, that is ioo million million times brighter. Each factor of 
ioo in brightness corresponds to 5 magnitudes, so 100 million million 
means a difference of 35 magnitudes, and the supernova would appear 
magnitude 16 — 35 = — 19 if it were at a distance of 10 parsecs — far 
brighter than the Full Moon. 

The third correction that ought to be made is for the amount of light 
from the supernova that is absorbed by dust in our Galaxy, and in the 
parent galaxy of the supernova. The absorption correction corresponding 
to our own Galaxy's dust can usually be made with adequate accuracy 
on the simple assumption that the dust forms a slab in our Galaxy, 
parallel to the Milky Way, through which astronomers point their 
telescopes towards supernovae in different directions, looking through 
different slanting thicknesses of dust. Astronomers have no information 
on the absorption in the parent galaxy, because they cannot tell whether 
the supernova is on the nearer or farther side of the galaxy. 

The latest result of these calculations on the absolute magnitude of 
Type I supernovae is that the average magnitude is -19.5 to -20, 
corresponding to a luminosity nearly ten billion times that of the Sun. 
This number is considerably in excess of the first estimates (in the 
1930s) of the luminosity of supernovae because at the time the scale of 
the Universe was considerably underestimated. 

Knowing now that the absolute magnitude of Type I supernovae is 
about - 19.5, we can attempt to calculate how far away such a supernova 
has to be pushed in order to appear magnitude -4.0 (Tycho's super- 
nova) or - 2.6 (Kepler's supernova). These calculations are somewhat 
handicapped by the correction for dust in our Galaxy. Since these two 
supernovae are within the galaxy, the assumption that the absorption 
lies in a slab is no longer relevant (the supernovae might be within the 
slab). We can, instead, use the fact that dust in the Galaxy not only 
absorbs starlight, it reddens it too, just as the Sun is reddened at sunset 
when its sunlight passes slantwise through the Earth's atmosphere. If 
we can obtain information about the color of the two early supernovae 


TYPES of supernovae: gathering the evidence 

and know what color (by present-day measurement of Type I super- 
novae in other galaxies) they might have appeared had there been no 
dust, then we can determine how much dust lies between Earth and the 
supernovae. We have to rely on subjective impressions by Tycho and 
Kepler of the color of the respective supernovae and the argument 
treads on quicksand at this point. 

However, all observers stressed that at its maximum Tycho's super- 
nova had the yellow color of Venus or Jupiter, becoming redder like 
Mars or Aldebaran after a month and whiter like Saturn at the same 
time thereafter. Kepler's supernova seems to have been even redder on 
the whole than Tycho's. On the basis of a discussion of these observa- 
tions, Minkowski concluded that the absorption of light from Tycho's 
and Kepler's supernovae by dust in our Galaxy was 2.1 and 3.3 magni- 
tudes respectively. Thus, had there been no dust, they would both have 
appeared at magnitude - 6, or 14 magnitudes fainter than if at the 
standard distance of 32 light years. They are both, therefore, some 
20,000 light years distant. We have to be wary of the large uncertainty 
in this estimate, caused mainly by the doubt about the reddening 

What about the light curves of the Type II supernovae? There is 
considerable diversity in these but after the maximum, Type lis typi- 
cally fade by a magnitude and a half and then almost halt their decline 
for 50 days. They then decline more rapidly at no particular rate, fading 
by two magnitudes in 60 to 120 days. Because Type lis are more 
diverse, because they are rarer than Type Is, and because few have been 
well observed, their absolute magnitude is not known with certainty, 
but there is a suspicion that they may be fainter than Type Is by a 
magnitude or two. 

Spot the difference 

The original distinction between Type I and Type II 
supernovae was made by Minkowski by looking at their spectra. By 1941 
he had taken spectra of 14 supernovae and found that nine were very 



similar in showing various distinctive bands of colors, whereas spectra 
of the five examples of the second type were featureless during the 
period of maximum brightness and then developed distinctive features 
because of hydrogen. The Type lis look most like common novae— 
their spectra have blue-shifted spectral lines, indicating that material is 
approaching Earth at speeds as high as 15,000 km/s at the time of the 
supernova outburst. Presumably these lines are caused by a shell of 
material ejected from the supernova, which of course on its near side 
will be approaching Earth. Type Is have less understandable spectra, 
but recently it has been established that they have the same basic cause, 
though the expansion speed of the shell of material reaches some 
20,000 km/s. 

The difference between Type I and Type II supernovae is not just 
the distinction between two engravings of the same postage stamp, a 
fascinating exercise with little meaning beyond itself. The clue to the 
significance of the difference is given by the kind of galaxies in which the 
two types of supernovae are found. 

Not all galaxies are alike. Once Hubble had established that galaxies 
were beyond the Milky Way (extragalactic) he went further. From the 
vast numbers of photographs which he accumulated, he established the 
existence of a series of galactic shapes, ranging from irregular con- 
glomerations of dust, gas and bright young stars, through spiral galaxies 
with arms in which the dust, gas and bright young stars concentrate, to 
structureless spherical or elliptical balls of old, red stars with scarcely a 
trace of dust and gas. Although Hubble himself warned against the 
implication that this sequence of galaxian forms represented the 
evolutionary life of a galaxy, it is now thought that there is some connec- 
tion between the appearance of a galaxy and the age of the stars within it. 
Only in spiral galaxies and in irregular galaxies is there still a supply of 
dust and gas to form young stars now. For some reason star formation in 
elliptical and spherical galaxies has long ceased. The massive bright 
young stars in elliptical galaxies have all ceased to shine and only stars 
like the Sun still visibly survive in them. 


TYPES of supernovae: gathering the evidence 

When astronomers examine the frequency with which supernovae 
occur in various kinds of galaxies, they find that no Type II supernova 
has ever occurred' in an elliptical galaxy, or even in the kinds of galaxies 
considered as intermediate in shape between the ellipticals and the 
spirals. Only in galaxies with clear spiral arms have Type lis ever been 
seen. Type Is on the other hand appear in all kinds of galaxies but favor 
the elliptical galaxies. The implication is that Type II supernovae are 
the consequence of the evolution of the more massive stars (with masses, 
say, of ten times the solar mass) whereas Type Is are caused by the 
evolution of stars typically of one solar mass (or less). 

While most astronomers are happy with this notion about the pre- 
cursors of Type lis, many have expressed qualms about accepting the 
contrary implication for Type Is. Some have proposed that these come 
from exploding white dwarf stars or from close double stars. Some have 
claimed that there must be a few very massive stars in elliptical galaxies, 
though not as many as in spirals, and that Type Is, like Type lis, are 
both formed from massive stars. The origin of the more frequent type of 
supernova, the Type Is, remains obscure. 


IX The making of a neutron star, 
what makes pulsars tick 

The stars, like measles, fade at last. 
Samuel Hoffenstein 


"o far we've seen how astronomers have demonstrated 
observationally that massive stars produce supernovae and that super- 
novae make neutron stars. Knowing that this occurs is not enough; 
astronomers want to know why. To understand why we have to look at 
the life history of stars in some detail. 

The life and death of stars 

The modern theory of stellar evolution goes like this. All 
over the Galaxy there are clouds of interstellar hydrogen, vague, tenuous, 
gaseous masses, called nebulae when they can be seen shining, which they 
do if they are near a bright star. For reasons which are not entirely clear, 
especially dense parts form within a nebula, and these dense parts have 
a sufficiently strong mutual gravitational attraction to make them fall 
together in larger and larger lumps called protostars. They are "about- 
to-be" stars. Gravity is the force which attracts these lumps to one 
another, just as the force of gravity on Earth pulls all objects down to the 
surface, drawing them inexorably towards the center. 

As the gas of which the protostar is made is packed closer and closer 
in towards its center, it begins to heat up, just as the air in a bicycle pump 
heats as it is compressed into the tire. You can feel the valve getting hot 



as you pump. Such heating causes the star to stop collapsing. It causes 
the pressure in the center of the star to increase, and this pressure 
shoulders the atoms of gas apart which in turn keep the outer layers 
from collapsing. The star becomes a finely balanced mechanism, 
tending to collapse into itself because of its own gravitational force, but 
prevented from doing so by the pressure of the gas inside. 

This balance comes about only when the gas in the center of the star 
reaches a temperature of many millions of degrees. In a star like the Sun, 
the center reaches a temperature of 15 million degrees centigrade and a 
density of about 160 times that of water. In the innermost three per cent 
of the volume of a star like the Sun, into which is packed two thirds of 
the mass of the star, the temperature and the density are so high that 
nuclear reactions take place. Over the whole star the hydrogen atoms 
from which the star is made have been split apart into their component 
bits by the force of the collisions between the atoms as a result of the 
high temperature ; the hydrogen atom is so simple that it produces only 
two pieces, an electron and a proton, its massive central nucleus. In the 
very center of the star the protons themselves have been forced so close 
together that there is a substantial chance that four protons will be able 
to stick together in a new arrangement called an alpha particle which is 
the central massive part, or nucleus of a helium atom. As the four 
hydrogen nuclei are converted into a helium nucleus, energy is released, 
mostly in the form of gamma rays, which are very energetic light- 
particles or photons, similar to light but with a million times more 

The gamma rays slowly diffuse up out of the star, being degraded on 
the way to a much larger number of lower energy photons which ulti- 
mately leave the surface of the star in the form mostly of light and 
infrared radiation. It is this light which leaves the star, travels through 
space, and permits stars to be seen. Indeed, sunlight is the result of 
gamma rays emitted from the center of the Sun millions of years ago, 
released by the nuclear reactions, and providing a fossil record of these. 

The energy production within a star can be thought of as brilliantly 



shining proof of Einstein's famous equation E = mc 2 — that is, mass and 
energy are equivalent, and one can be converted into the other given the 
right conditions. Each time four protons are converted into a helium 
nucleus, 0.7 percent of the mass of the four individual protons is con- 
verted into radiative energy. Although each nuclear event converts just 
a small amount of mass to energy, in total a star like the Sun radiates 
four million tons of its mass away every second of its life. 

At this rate of consumption even a mass as large as that of the Sun 
must be appreciably diminished in the course of time and less hydrogen 
is available to sustain the rate of energy production. The star becomes 
unbalanced, unable to support itself against its own gravitational 
attraction. The center of the star begins to contract and the outer parts 
begin to expand, producing what is technically known as a red giant. 

In a sunlike star, this occurs at an age of ten billion years. The con- 
version of the four hydrogen protons to a helium nucleus still takes place, 
not in the center but in a shell around the center of the red giant. The 
center of the star, which is by now largely helium, releases energy first by 
contracting, but as it contracts, if it is large enough, it eventually becomes 
hot and dense enough for a further kind of nuclear reaction to begin. 
The star begins to convert the helium nuclei into carbon nuclei by what 
is known as the triple-alpha process, because it involves the uniting of 
three helium nuclei or alpha particles to make a carbon nucleus. 

This further conversion of energy in the triple-alpha process halts 
the contraction of the center of the star for a while, but not for long. 
Because the triple-alpha process is a much less efficient means of 
generating energy than the conversion of hydrogen to helium, the star 
remains a red giant for only a comparatively short time. The star's 
central regions contract even more, and if the star is massive enough, 
the contraction causes the center to heat up enough for yet further 
nuclear reactions to occur. In these, helium nuclei are successively added 
to carbon nuclei to form heavier nuclei like oxygen, neon, magnesium 
and silicon, possibly up to iron nuclei. But less and less energy is avail- 
able from these reactions and they only briefly postpone the ultimate 



collapse of the center of the star to a very dense state, known as degenerate 
matter, when the red giant star turns into a white dwarf. 

Because it has collapsed so much, the density of a white dwarf is very 
high: about one million grams per cubic centimeter, so that a block 
one centimeter on the side — roughly the size of a sugar lump — weighs 
one ton. Left to itself, the white dwarf gradually cools off, rapidly at 
first from white hot to yellow hot, but then more slowly to red hot. As it 
becomes redder, it fades until, ultimately, it disappears from view. At the 
last stage, it is known as a black dwarf. 

In 1932 the Indian-born American astrophysicist S. Chandrasekhar 
proved a remarkable theorem about white dwarfs, showing that no 
white dwarf could exceed a mass of about one and a half times the mass 
of the Sun. A star of larger mass which attempted to become a white 
dwarf would not be able to hold itself up against the force of gravity and 
would have to turn into something other than a white dwarf. This 
maximum mass for a white dwarf is known as the Chandrasekhar Limit. 
The Chandrasekhar Limit lies between 1.44 solar masses and 1.76 solar 
masses depending on precisely which nuclei the star is made of. If it is 
composed of helium nuclei, the maximum possible mass is 1.44 solar 
masses, and if of iron nuclei, 1.76 solar masses. Yet the number of white 
dwarfs known is too large for them to have evolved only from stars which 
are less than the Chandrasekhar limiting mass. (Near the Sun there are 
about five white dwarfs in every cube of space whose side is 30 light 
years.) Somehow, stars which are heavier than the Chandrasekhar 
limiting mass must lose material before they can become white dwarfs. 

How weighty stars lose mass 

Stars can lose material at two stages in their evolution. The 
first is while they are red giants. Because a red giant is very extended, 
its mass is spread over a large volume and it has a low surface gravity. 
Bits of the atmosphere can, relatively easily, be thrown back into space 
by various storms, flares and winds from the star's surface. Then more 


-.'•' " : ."'^.• ; -" ; V/«y' : "' : i -;'. 

A more massive star than the 
Sun, rotating sedately 

. . . becomes a red giant star 
with carbon core . . . 

dramatic ejection of material into space occurs as a red giant attempts to 
become a white dwarf, during which it puffs off a layer of material from 
its surface. As this sphere of material slowly expands away from the star, 
it becomes so huge that it can be photographed by astronomers, who call 
it a planetary nebula. This is because when seen through a small tele- 
scope, it has the appearance of a flat disk like the planets Uranus and 

A typical planetary nebula is comparable in size to the solar system, 
but the largest can be a light year or more in diameter. Its mass of around 
two tenths that of the Sun expands into space at a speed of about 
20 kilometers per second. In a time of several thousand years, the nebula 
disperses into space. Perhaps, in order to become a white dwarf, a 
massive red giant has to go through the stage of ejecting a planetary 
nebula several times in order to bring its mass below the Chandrasekhar 
Limit. Through this process, stars as large as four times the solar mass 
can become white dwarfs by ejecting planetary nebulae. 

Stars larger than about four solar masses cannot cope in this way with 
the problem of turning quietly into white dwarfs — the forces on them 
are too extreme. In this case, what exactly does occur is not clear, but it 
is some kind of catastrophe. Because of the successive "burning" of 
heavier and heavier nuclear fuels, the central part of the star is made of a 



series of concentric shells of different nuclei, like an onion, the inner 
shells containing the heavier nuclei. One of the layers, of carbon or 
oxygen, explodes and drives the layers outside it off the star. The core 
inside the exploding layer implodes ("exploding inwards," like a frac- 
tured light bulb), and shrinks so that its density increases beyond even 

... in a supernova explo- 
sion the core collapses, 
emitting neutrinos (v), 
and ejecting the outer 
layers of the star . . . 

. . . which become a 
shell-like supernova 
remnant enclosing a 
rapidly rotating pulsar 


the degenerate density of a white dwarf. It becomes, in fact, a neutron 
star, since electrons in the core of the star are forced into the protons 
present in the carbon nuclei. The negative electric charge of the electrons 
cancels out the positive charge of the protons, giving electrically neutral 
particles called neutrons. 

The material outside the degenerate carbon core must all be blasted 
off the star in the ferocious explosion. An explosion of this kind is what 
astronomers see as a supernova. The explosion ejects most of the mass 
of the precursor star. Neutron stars are subject to a limit on their mass 
similar to the Chandrasekhar limit on the mass of a white dwarf. There 
is some discussion about the size of the critical mass of a neutron star, 
but C. E. Rhoades and R. Ruffini of Princeton University have calcu- 
lated that the critical mass certainly cannot exceed 3.2 times the mass of 
the Sun. All neutron stars have masses below this figure, probably 
below 2.5 solar masses. A typical neutron star has the same mass as the 
Sun. In the explosion of a five solar mass star, therefore, three or four 
solar masses must be ejected if a neutron star is to be the result. 

The density of neutron stars is extremely great, at least ten million 
million grams per cubic centimeter, so that a cube the size of a sugar 
lump would weigh ten million tons, comparable with all the materials 
used to build a city the size of London or New York. This is because 
material to the amount of about one solar mass is packed into a star 
whose diameter may be only 20 kilometers. 

How spinning stars speed up 

A curious thing happens to the rotational period of a star 
which collapses into a smaller volume. All stars rotate to a greater or 
lesser extent. The Sun, for example, spins once on its axis every 25 days 
9 hours, just as the Earth spins on its axis in one day. Stars more massive 
than the Sun generally rotate faster, and periods as short as half a day 
are known for some stars. When stars expand and become red giants, 
their period of rotation lengthens : they slow down. During their subse- 
quent collapse to white dwarfs or neutron stars, they speed up again. 



The reason for this is a physical law called the conservation of angular 
momentum, which says that in this kind of situation the cross sectional 
area of a star divided by its rotational period will remain roughly a 
constant. When the star expands in the red giant stage its cross sectional 
area gets larger and therefore its period will become proportionately 
greater — the star slows down. When the star subsequently shrinks, its 
cross sectional area becomes smaller, its period is reduced in propor- 
tion — and the star speeds up. 

Ice skaters exploit the law of the conservation of angular momentum 
when they pirouette on the ice with arms outstretched, giving them- 
selves a large cross sectional area, and then rapidly bring their arms to 
their sides. This decreases their cross sectional area and thus decreases 
their period of rotation; in other words, they spin faster. Consider a star 
like the Sun with a cross sectional area of 600 billion square miles and a 
period of about one month (two million seconds). If it were possible to 
collapse the Sun suddenly to the size of a neutron star, the cross sec- 
tional area would decrease by a factor of a billion so that the period too 
would have to decrease by a factor of a billion, to about a five hundredth 
of a second. In a real supernova explosion not all the star collapses to a 
neutron star and the parts which are ejected carry off some of the 
angular speed of rotation so that in practice neutron stars rotate more 
slowly than this — their periods are several times longer although still 
almost incredibly fast. The Crab Nebula pulsar, when it started to 
rotate, must have been spinning about twice as fast as it is now, with a 
period of about a sixtieth of a second. 

The rotation of heavy objects makes a very good repetitive clock. 
After all, the Earth was used as a clock until very recently when atomic 
clocks revealed the small irregularities in its rotation. The regular 
rhythm of a pulsar, and the Crab Nebula pulsar's rate of repetition of a 
few tenths of a second, led Thomas Gold to suggest that pulsars were 
rotating neutron stars. In some way, he suggested, a pulsar emitted a 
beam of light like a lighthouse, and this beam, as it turned towards the 
Earth with each rotation, produced the pulses observed by radio and 



optical astronomers. When he found that the Crab pulsar was slowing 
down, feeding its pulses from its rotational energy, he became con- 
vinced that his idea was correct. 

How pulsars may shine 

It is not quite enough to say that pulsars are like lighthouses, 
beaming their light and radio rays into space as they spin around. How 
actually do they make the beam ? The answer is unclear but tied up with 
an immense magnetic field at the surface of a pulsar. All stars have 
magnetic fields to a greater or lesser degree and when a star produces a 
pulsar its magnetic field is caught fast and throttled, squeezed into the 
smaller cross section of the neutron star. When this happens the strength 
of the magnetic field increases enormously, by a factor of millions. 

Pulsar astronomers assume that the magnetic field of the pulsar does 
not lie exactly along the polar axis of the pulsar. As in the Earth, whose 
magnetic axis leaves its surface in Canada, 15 of latitude from the 
North Pole, the magnetic field of a pulsar may lie off the pole of rotation, 
possibly even at the pulsar's equator. The pulsar is therefore a powerful 
electric generator, spinning a magnetic field lying across a rotor, and an 
electric current pours from the surface of the neutron star. Caught in the 
magnetic field the electrons which make up the current stream out of 
the magnetic poles radiating as they do so. The beam of radio waves 
which they emit points out of the magnetic axis of the pulsar and as the 
magnetic axis sweeps in the direction of the Earth we perceive a pulse. 
Possibly we are in a position to see a diminished amount of radiation 
from the magnetic pole on the far side of the pulsar and then we can see 
two pulses : many pulsars, including the Crab, show two pulses during 
every rotation. 

Perhaps not all pulsars shine in this way. Perhaps in some the beam 
of radiation is produced by electrons bunched, not at the magnetic 
poles, but off the surface of the neutron star near to the place where the 
magnetic field is moving at speeds close to the speed of light. Then the 



radiation is caused not by electrons beaming along the magnetic field 
but along the direction in which they are moving, at right angles to the 
magnetic field. 

Whatever the cause, investigation of the way pulsars shine has 
stimulated research into the fundamental physics of fast-moving strong 
magnetic fields and particles. Only in the cosmic laboratory has it been 
possible to see such an experiment in operation. 


Although pulsars are very good clocks, rotating very regu- 
larly apart from a gradual slowdown, most pulsars, particularly the two 
fastest, the Crab Nebula pulsar and the pulsar in the Vela supernova 
remnant, do, when examined closely enough, show irregularities. At the 
end of September 1969, for example, the period of the Crab pulsar 
suddenly decreased by a third of a billionth of a second. There have 
been three much larger jumps in the Vela pulsar's period, each of 
200 billionths of a second. The reason for these sudden so-called 
"glitches" seems to be that neutron stars have a crust, just as the Earth 
does. Because they are rotating so fast the crust is not precisely spherical 
but is flat at its poles and bulges at the equator, due to the centrifugal 
force there, just as the Earth is tangerine shaped. As the pulsar slows 
down the centrifugal force becomes less, but the crust takes the strain 
for a time even though the equatorial bulge of the pulsar tries to fall. 
The crust cannot survive the stress indefinitely. It breaks and suddenly 
drops. By the law of the conservation of angular momentum previously 
mentioned, the pulsar suddenly spins faster. 

The amount by which the crust drops is minute — it is measured in 
millimeters ! But this is enough to have a noticeable effect on the pulsar's 
period just because it is usually so regular. 

Astronomers call the sudden changes in the neutron star "star- 
quakes." The energy released in a starquake is enormous and flows out 
from the star into the space surrounding it. After the 1969 starquake on 



the Crab pulsar a wave of energy was seen to flow outwards, rippling 
through the center of the Crab Nebula. This, apparently, is the cause of 
the waves of activity which Baade noted and was one of the clues which 
convinced him correctly in 1945 that he had identified the Crab Nebula 
supernova's stellar remains. 


X Neutrino astronomy: 
the ultimate cause of 

O dark dark dark. 
They all go into the dark, 
The vacant interstellar spaces, 
The vacant into the vacant. 
T. S. Eliot 


•ost things in science which we know about have 
been discovered by somebody. Yet it is possible to talk about some things 
actually being invented, in the sense that a theoretical scientist saw that 
logically they had to exist before they had been discovered. This is so in 
the case of the neutrino, which was invented by Wolfgang Pauli in 1930. 
The story of the neutrino is also a good example of the way in which the 
behavior of sub-microscopic particles, individually all but undetectable, 
can have devastating effects on a large scale. For, apparently, neutrinos 
can actually cause a supernova. 

Pauli had to invent the neutrino because of a fundamental law of 
science: energy can be changed from one form to another, but it never 
appears or disappears. Atomic scientists in the 1920s were worried that 
this law was apparently being broken in a certain kind of nuclear trans- 
formation called beta-decay, which is, essentially, the way in which a 
neutron spontaneously decomposes into a proton and an electron. A 
proton is positively charged and an electron is negatively charged, while 
a neutron has no charge at all. Consequently, the result of a neutron 
suffering beta-decay is still no net electric charge, since the charges on 
the electron and the proton cancel each other. Therefore, electric charge 
did not appear or disappear in beta-decay: in the jargon of nuclear 



scientists, it was conserved. But although the charge was conserved, the 
problem was that the total energy of the neutron alone was greater than 
that of the pieces after beta-decay. Energy was not being conserved : it 
seemed to disappear. 

To get over this difficulty, Pauli invented a particle which could have 
no charge, but which would balance out the energy equation. A small 
fraction of the energy in beta-decay, he said, was carried off by this 
imagined particle which no one had ever detected. Because the energy 
carried away was sometimes very small, the neutrino had to have a very 
small mass — even zero! Pauli was aware that he was dangerously near 
sophistry. He thought that no one would ever detect the neutrino and 
told the astronomer Walter Baade, "Today I have done the worst thing 
for a theoretical physicist. I have invented something which can never be 
detected experimentally." Pauli originally called his imaginary particles 
"neutrons." But they were different from what we now call neutrons, 
which were not actually discovered until 1932. Enrico Fermi, with 
exasperation and gestures to match, explained the difference to an 
audience of slow physicists at a conference in 1933: "The neutrons 
discovered by Chadwick are big. Pauli's neutrons were small. They 
should be called neutrinos." The -ino ending in Italian is a diminutive, 
like bambino, and the name stuck. 

Fermi worked out that the chances of a given neutrino reacting with 
anything were very small. If a neutrino travels at the speed of light 
through a 3000-light-year-thick slab of matter with the density of water 
(the average density of the Sun), it has only a 50-50 chance of reacting 
with a proton. 

Nonetheless, neutrinos are given off in large numbers from nuclear 
reactors : ten trillion every second pass through a square centimeter near 
the reactor. Although each one has only a small chance of being detected 
in experiments, there are so many available that just a few actually are. 
Pauli was too pessimistic — the existence of neutrinos has been con- 

Now, stars like the Sun are vast nuclear reactors : the Sun creates in 



its center vast numbers of neutrinos every second. Because the Sun's 
radius is only 2 light seconds, very small compared with 3000 light years, 
nearly all the neutrinos created by the Sun dash unheedingly out of the 
Sun's surface and diffuse through space. By the time they reach Earth, 
the number passing through each square centimeter per second has been 
calculated to be 65 billion : a much smaller number than can be made in a 
terrestrial nuclear reactor, but still just enough to give a few detectable 

Detecting the undetectable 

How can scientists detect a particle which will hardly 
interact at all with matter, passing right through the Earth without being 
affected? In 1946 B. Pontecorvo, an Italian immigrant to the U.S. who 
was later to defect to the U.S.S.R., suggested a method by which neu- 
trinos might be found. Pontecorvo's idea is now being realized in an 
experiment performed under the leadership of Raymond Davis of 
Brookhaven National Laboratory. 

The detector consists simply of a tank containing 610 tons of tetra- 
chloroethylene, a fluid used in the drycleaning trade. Each molecule of 
tetrachloroethylene consists of four chlorine atoms coupled to two carbon 
atoms. Approximately once per week, one neutrino out of the expected 
120 thousand million million million reaching the tank from the Sun 
interacts with a neutron in one of the chlorine atoms in the tank. The 
result is that the chlorine atom, which previously contained 17 protons 
and 20 neutrons, becomes an argon atom containing 18 protons and 
19 neutrons. 

After a while, the tank is bubbled for a day with helium which 
sweeps up the scores of argon atoms created by the solar neutrinos. 
They are recovered by freezing the argon out over charcoal. The created 
argon atoms are not stable : they suffer beta-decay on a time scale of a 
month and eject electrons which can be detected electronically. The 
strength of the signal detected in this way depends on the number of 
argon atoms which have been created in the tank since the last run. 



In practice there are spurious events occurring in the tank, caused 
by the radioactivity in nearby rocks and by cosmic rays. The cosmic 
ray contamination has been minimized by constructing the tank one 
mile below the surface in a gold mine in S. Dakota, so that most of the 
cosmic rays are absorbed by the mile of rock. Nevertheless, the presence 
of the spurious events has masked events caused by solar neutrinos. 

Before Davis began his experiment, calculations based on the assumed 
internal temperature of the Sun gave a value for the number of neutrinos 
to be expected. Expressed in terms of Solar Neutrino Units, the value 
was 6 snu. 

What SNU? 

The actual results from Davis's experiment came as the 
biggest shock to astrophysicists for many years. As his experimental 
procedures became more refined, it became clear that the neutrino rate 
was actually much, much lower than predicted, with at most a level of 
i snu, around the level of the spurious events. In fact, it can be said that 
no solar neutrinos have been detected with any confidence at all. This 
moved Willy Fowler, one of the leading astronomers in this field, to 
remark at a conference "What snu?" 

During 1976, somewhat higher rates were detected. Davis himself put 
this down not to a change in the Sun itself, but to the uncertainties of his 
apparatus. Even so, there remains the outstanding problem — why so 
few neutrinos ? 

Neutrinos from supernovae ? 

The Sun is a weak source of neutrinos compared with a 

Even before it forms a neutron star, the precursor of a supernova 
creates neutrinos in abundance by two main methods. In the first 
method, radiation creates matter. It does so in a beautifully symmetric 



way. To every kind of particle of matter there corresponds a kind of 
particle of antimatter, and, if it is energetic enough, radiation can 
produce matter- antimatter pairs of particles. A gamma ray produced 
at the center of a massive star is energetic enough to create an electron 
and its antiparticle, a positron. These can recombine and produce a pair 
of neutrinos. In the second method, an electron is captured by a proton 
in a nucleus with the emission of a neutrino, and the resulting neutron 
decays back to an electron, a proton and a neutrino. The neutrinos run 
off with a fraction of the energy, but the original nucleus still remains to 
suffer again this process of attrition. This is called the Urea process after 
a casino in Rio de Janeiro, where the customer loses little by little. 

The energy carried away by neutrinos from the center of a massive 
star is the very cause of the supernova explosion itself. Energy trans- 
formed into speeding neutrinos is lost from the star virtually instan- 
taneously, whereas energy transformed into radiation jostles its way out 
of the star and helps to support the star against its own gravitational pull. 
The more energy lost from the center of the star as neutrinos, the less 
support the released energy gives to the star. When support drops too 
far, collapse becomes inevitable. 

Initially, about half of the particles in the center of the star are 
neutrons; the other half are protons. Both neutrons and protons swim 
in a sea of electrons. The implosion forces the protons to swallow the 
electrons and make neutrons: the star center becomes a neutron star. 
Each creation of a neutron liberates one neutrino, increasing the neutrino 
output still further. 

Is there a possibility of detecting neutrinos from supernovae ? There's 
a bare possibility that they may have already been detected. Davis has 
performed more than 30 runs in his underground neutrino observatory, 
and in Run 27, performed in late 1972, a significant number of neutrinos 
was detected. There has been speculation that they were the result of a 
flash of neutrinos from a distant supernova, whose light went unrecorded 
because it was absorbed by dust in our Galaxy. However, there was 
another apparent increase in the neutrino detection rate in 1975-76 and 



two supernovae in three years seems too many. Perhaps these flashes are 
rare cosmic ray events, or experimental glitches, rather than supernovae. 
Davis's difficult experiment is giving tantalizingly equivocal results. 

Is the experiment worth pursuing ? It gives the first possible glimpse 
into the interior of the Sun and supernovae and the failure to observe 
the expected number of neutrinos has revealed astronomers' ignorance. 
Since the amount of material at and above the surface of a star — the only 
material visible to astronomers before the neutrino experiment — is just 
a few million millionths of the mass of the star and astronomers' idea of 
the structure of the remainder is educated guesswork, it is not surprising 
that calculations of the expected number of neutrinos from the Sun were 
wrong. The deficiency has inspired new efforts to understand better the 
interior of the Sun and other stars, though the American astronomer 
E. E. Salpeter's comment is still true, that "at the present time, we 
neither have a positive identification of solar neutrinos nor the morbid 
satisfaction of predicting a scandal in stellar evolution theory." More 
than this, the stimulus which the experiment has given to neutrino 
astronomy has meant that the properties of the neutrino have been 
subjected to particularly close scrutiny. Its interactions are so rare that 
terrestrial experiments have difficulty in detecting neutrinos at all. But 
in a supernova explosion so many neutrinos are released and the star is 
so dense that sufficient neutrinos may be absorbed to have a significant 
effect on the explosion of the star. The study of supernovae — cataclysms 
on the grandest scale — is revealing properties of the neutrino — nature 
on the tiniest scale. 

1 60 

XI Creation of the elements: 

Man, the supernova remnant 

Act first, this Earth, a stage so gloomed with woe 
You all but sicken at the shifting scenes. 
And yet be patient. Our Playwright may show 
In some fifth Act what this wild drama means. 



.n some of the most rugged mountainous country of 
New South Wales, along the Turon River and in its hinterland, are the 
ghost towns of the Australian gold rush of the 1870s. A few people are 
left in Hill End, but of the town of Tambaroora nothing remains ex- 
cept Golden Gully, a canyon dug by miners in their thousands. Tens of 
millions of tons of dirt were dug from here, washed and scrutinized for 
the glint of a metal : precious because both beautiful and rare, and made 
rare by the rarity of supernovae. 

Abundance of the elements 

Lured to inhospitable terrain because of the incredible 
richness of the strike, those miners who were luckier than most took 
from the goldfield a total of just 20 tons of gold, on average one gram 
for every ton of dirt dug over. Modern gold mines operate profitably 
when there is a worthwhile concentration of gold of about 20 grams 
(about an ounce), per ton, in contrast with the average concentration of 
gold in the surface of the Earth of approximately one thousandth of a 
gram per ton. 


Golden Gully, Tambaroora, New South Wales. Millions of tons of dirt were 
dug by pick and shovel from this gully to collect gold, grain by grain. The rarity 
of gold is linked to its formation deep inside massive stars. Because supernovae, 
scattering their heavy elements throughout space, are rare, gold is rare too. 



What of the concentration of gold averaged over the entire mass of the 
Earth ? There is, of course, no direct evidence since the central regions 
of the Earth are inaccessible. However, it is a speculation commonly held 
by astronomers that meteorites, stones and rocks which have fallen from 
space to the Earth, represent the remains of a defunct planet as solid as 
the Earth or Moon, so that the abundance of the elements in meteorites 
may be like the abundances in the interior of the Earth. Gold is ioo to 
200 times more abund ant in m eteorites than in the surface of the Earth, 
so is presumably similarly more abundant in the Earth's center. 

It is not difficult to guess why this should be. The Earth's interior is 
hot, because of heating by radioactive materials, and partly perhaps 
because of its contraction under the force of gravity. In some ways it 
resembles an ore-smelting furnace, melting rocks so that the lighter slag 
rises to the surface to make the Earth's crust, while the metals such as 
iron, nickel and gold fall towards the Earth's center. 

Determining the abundance of gold in stars directly is not possible. 
Only one clear spectral line in the Sun's spectrum caused by atoms of 
gold has been found, proving gold's existence there, but since the 
mechanism in the gold atom that causes the spectral line has not been 
studied well enough, no accurate estimate of the amount of gold 
required to form the line can be inferred. Its concentration in the Sun 
has been estimated by assuming that, since nickel and gold have similar 
properties, the ratio of their concentration in meteorites is the same as in 
the Sun. Because nickel gives rise to many well-studied spectral lines in 
the solar spectrum, its abundance can be easily measured. 

Approximately one milligram in every ton of the Sun is gold. 

Why is gold rare ? Why is iron on the other hand relatively common ? 
Why is it that in spite of the great diversity of astronomical objects whose 
composition has been studied — the Earth, meteorites, the Sun, most 
stars — the relative abundances of the elements in all these bodies are 
surprisingly similar, and the differences are readily explainable by some 
plausible guesses and the histories of the bodies ? 

Clearly there is some common astronomical explanation for the origin 



The valleys on this graph of 0.03% of the known spectrum of the 
Sun represent chunks of color "bitten" from sunlight by atoms, 
mostly of iron, in the Sun's atmosphere. Peaks represent colors 
which have passed relatively freely out of the Sun. The nibble 
arrowed is the only evidence for the existence of gold in them there 
solar hills. It hints at the wave of supernovae which manufactured 
the gold, and enriched the pre-solar nebula from which the Sun 

of the elements in all these celestial objects, and supernovae play a 
crucial role. 

The alpha-beta-gamma theory 

Modern discussions have a two-pronged attack on the 
creation of the elements. The first starts, as the Universe started, with 
the Big Bang. This theory of the origin of the elements was originally 
called the alpha-beta-gamma theory, in part because the elements are 



supposed to be formed in sequence like the start of the Greek alphabet, 
and in part because the theory was proposed in detailed form by 
Ralph A. Alpher, Hans Bethe and George Gamow in 1948. (Bethe's 
part in creating the theory was small — Gamow said that he invited 
Bethe to be a coauthor because the pun appealed to Gamow's sense of 
humor.) Gamow called the material of the Big Bang ylem: he envisaged 
it as a gas made of neutrons, although modern authors see it as a more 
complicated mixture. When the theory was first put forward it was to 
account for the formation of all the elements from this basic ylem. As we 
shall see, however, it would complete only the first stage of the process. 

The neutrons in ylem changed relatively slowly into protons and 
electrons as the Universe got under way. Some of the protons thus 
formed captured neutrons to make more complicated nuclei. Some of 
these nuclei would change by beta-decay (emission of an electron) and 
some would gather further neutrons to become more complex nuclei. 
All the element building in the Universe, according to the alpha-beta- 
gamma theory, occurred in the first two hours that the Universe existed. 
As the Universe cooled the nuclei would capture the free electrons to 
become atoms. 

One feature of the abundance of the elements which this theory 
explains well is the fact that nuclei whose ability to capture neutrons is 
low are more common than those whose ability is high. Think of nuclei 
as a large intake of graduates into a big organization in which promotion 
is by merit. Bright graduates (nuclei with high neutron-capture ability) 
are susceptible to promotions; they take advantage of random oppor- 
tunities (neutrons) as they occur and are promoted faster than their 
fellows (they form more complex nuclei). The duller graduates' ability is 
lower, they stay in their career grades longer and consequently there are 
more of them than of the bright graduates. 

Among the elements certain nuclei have exceptionally low ability to 
capture neutrons — these occur at the so-called magic neutron numbers 
50, 82 and 126. Elements occurring at these numbers, like lead, are 
exceptionally common. 


Big bang 

Hydrogen & helium 

_». Interstellar material 



White dwarfs 

Neutron stars 

Black holes 

= route described in this book 

Chemical evolution -of the Galaxy. The flow is from the Big Bang which created 
hydrogen and helium from which stars form. Gas is recycled back into the 
interstellar medium via planetary nebulae or supernova remnants, or ejected by 
low surface gravity supergiants, but everything ultimately ends as white dwarf, 
neutron star or black hole. 



The alpha-beta-gamma theory suffers, however, from a fatal flaw. It 
requires that the nuclei build up from hydrogen by adding one neutron 
at a time, with the neutron changing to a proton at appropriate stages. 
The flaw is that at two vital stages, the nuclei created cannot exist for 
more than a minuscule fraction of a second (a thousand, million, million, 
millionth of a second !) after which time they release the neutron that 
they have just captured and return to a helium nucleus again. They do 
not exist for long enough for the next step to occur. The break in the 
chain occurs just after the formation of helium, about two minutes after 
the start of the Big Bang. Thus the alpha-beta-gamma process cannot 
proceed past helium in making the elements. No further heavier ele- 
ments can be made. If the alpha-beta-gamma process of element build- 
ing were all that could occur, the Universe would consist solely of 
hydrogen and helium and there would be no carbon, no nitrogen or 
oxygen, no paper to make this book, no writer to write it, no reader to 
read it. 

Creating elements inside stars 

There must have been other sites in the Universe, apart 
from its beginning, at which the heavier elements were created. When 
the process by which nuclear energy became starlight was discovered as 
a result of the work by Bethe and von Weizsacker in 1939, it was realized 
that the same processes would change the composition of the stars and 
create new elements. 

The first observational evidence that the stars create elements was 
Paul Merrill's discovery in 1952 of spectral lines of the element tech- 
netium in red giant stars. Technetium is unstable and lasts at most for a 
few million years. As the red giant stars were known to be older than 
this, clearly technetium could not have been in these stars since they 
were formed ; it must have been made there. Other stars were discovered 
which, like red giants, had developed sufficiently in their evolutionary 
life to show an excess of carbon or nitrogen caused by helium burning in 
the so-called triple-alpha process. 



The time was ripe for a detailed examination of the formation of the 
elements in stars. 

The problem engaged the attention of Fred Hoyle in 1946 because of 
his advocacy of the Steady State Theory of the Universe in which there 
was no Big Bang and therefore no cosmological element creation. Hoyle 
was faced with the existence of the wide variety of elements which he 
had to explain in other ways, and one of the main successes of the 
Steady State Theory was to stimulate this research, although the theory 
seems to have since lost credibility as a cosmology. The specific pro- 
cesses which form the elements in stars were detailed in a foundation- 
laying paper in 1957 by Geoffrey Burbidge, Margaret Burbidge, Willy 
Fowler and Fred Hoyle, known as the B 2 FH (B-squared, F, H) paper. 
B 2 FH supposed that the first stars consisted principally of hydrogen. 
Most stars visible now are in the process of converting that hydrogen 
to helium, releasing energy which can be seen as starlight. 

This process creates helium from hydrogen. As stars age, they 
"burn" some of the helium which they have created to produce carbon, 
oxygen, neon and magnesium. The interior of the star may be mixed, 
stirred up by clouds of hot material billowing from the star's center 
towards its surface by the force of convection. Depending on whether 
mixing occurs or not, the carbon and oxygen can capture protons 
(hydrogen nuclei) or alpha particles (helium nuclei) to make either 
proton-rich nuclei or the elements magnesium, silicon, sulfur, argon and 

B 2 FH identified five further processes occurring in or on the stars. 
The first, which they named the ^-process, occurs somewhere in some 
kind of star yet to be satisfactorily identified. When the mixture of 
elements formed in the previous processes cooks at a high temperature, 
the protons in the mixture absorb electrons and release them at equal 
rates in a situation of equilibrium (hence the name ^-process). This 
creates those abundant elements such as iron, nickel, chromium and 
cobalt which are known collectively as the iron peak. 

Up to this point the element building process has been relatively easy : 



the creation of heavier elements has released energy. In a sense, the 
nuclei of these elements want to be formed (in the same sense that a 
heated object wants to cool by radiating energy). Beyond the iron peak 
however, the processes creating heavier elements have to have a supply 
of energy available to do so. Such a supply is available in a supernova 
explosion. Just before the supernova, the precursor star has built up a 
supply of middleweight elements by burning helium and carbon. 
Something happens and the precursor becomes unstable: the precise 
way in which the supernova occurs may not be important from the point 
of view of the creation of heavy elements. What matters is the speed at 
which the process occurs, so that neutrons are added to the middleweight 
nuclei sufficiently quickly that the successive nuclei, though unstable, 
do not have time to eject electrons and turn into something else. Because 
it is rapid, this process is called the r-process. (The ^-process is one in 
which successive neutrons are slowly added to middleweight nuclei 
which do have time to decay before the next neutron comes along.) 

The r-process forms most of each of the following elements: selenium, 
bromine, krypton, rubidium, tellurium, iodine, xenon, europium, 
gadolinium, terbium, dysprosium, holmium, erbium, thulium, ytter- 
bium, lutecium, rhenium, osmium, iridium, platinum, gold and 
uranium. The reader may find the names of some of these elements 
unfamiliar. Others he will know are precious because they are un- 
common. These are consequences of the infrequency of supernovae. 

A process called the />-process takes place when protons are added to 
nuclei formed by the r and 5 processes, but the /(-process creates only a 
minority of the atoms of any one element. It may occur as the supernova 
ejects its outer layers into space, these layers containing unburned 
hydrogen and hence abundant protons, as well as r and 5 process nuclei. 

Scattering the elements 

The elements were made in stars. How do they come to be 
found in the Earth ? 

The elements created in stars, including those created in supernovae, 



are thrown into interstellar space by supernovae themselves. The outer 
part of a supernova — perhaps most of it — is ejected into space by the 
force of the supernova explosion; this contains about one Earth mass of 
r-process elements. For a while — perhaps up to 100,000 years — the 
supernova remnant is visible in optical or radio telescopes but it soon 
merges with the general interstellar material, mixed up with the rest. 
New stars condense from the interstellar material, shining first as large 
cool stars, not visible to optical astronomers but emitting enough infra- 
red radiation to be detected by specialized infrared detecting telescopes. 
Such stars spin fast. Having contracted from a much larger slowly 
rotating interstellar cloud, new stars spin up as they contract, just as a 
neutron star spins up as it is exploded from the center of a supernova. 
Nonetheless, because it is an observed fact that most stars spin slowly, 
newly formed stars must shed their angular speed by creating planetary 
systems such as ours. Virtually the whole of the rotational energy of the 
solar system lies in the massive planet Jupiter, which orbits the Sun once 
every 12 years. But if Jupiter, and the other planets, had not been ejected 
from the Sun, its rotational period would have been less than about 
three hours instead of its present period of just over 25 days. The newly 
formed Sun therefore shed some of its material in what is known as the 
solar nebula to make the solar system; and that is how the r-process 
elements, like gold, traveled from a supernova in the distant past and 
came to be found in the Earth, to be sought and dug up by the miners 
of the gold rushes. 

Cosmological clocks 

How distant a past was it ? The study of the age of the 
elements is somewhat grandly called nucleocosmochronology, and it 
exploits the fact that some of the atomic nuclei created in the r-process 
are not stable — they spontaneously decay into something else, perhaps 
on a very long timescale indeed. For example, the nucleus of the iodine 
atom exists in two long-lived forms, one of which is very stable and lasts 



indefinitely, while the other decays into a form of xenon gas on a time 
scale of 17 million years. Both kinds of iodine are produced in approxi- 
mately equal amounts by the r-process in supernovae, but because one, 
so-called iodine- 129 changes to xenon- 129, the other, iodine- 127 
eventually predominates. 

While all the r-process elements are floating about in some inter- 
stellar cloud, the xenon- 1 29 produced by the iodine- 129 dissipates into 
space. But after the r-process elements have condensed into planets and 
meteorites, the xenon- 129 produced by the decay of iodine- 129 is 
trapped in the rock containing the iodine. The amount of xenon-129 in 
the rock thus tells how much iodine- 129 has decayed since the rock 

We know that there were equal amounts of iodine- 127 and iodine- 129 
when the r-process which formed them took place; we know how much 
iodine-129 has converted into xenon-129 while the elements were in 
interstellar space. Because we know how long this takes, we now know 
the time that the r-process elements were in space before the rock 
solidified : some 200 to 600 million years (say half a billion years, in round 

We have measured the time between the supernovae which formed the 
r-process elements which condensed into the solar system, and the 
formation of the solar system itself. 

Is this a reasonable time ? Stars form in the spiral arms of the Galaxy. 
There the dust and gas between existing stars is compressed enabling 
new stars to condense from it. Spiral arms swirl around the Galaxy, 
causing gas at a given place in the Galaxy to be compressed repeatedly, 
the time interval between compressions being measured in hundreds of 
millions of years, the rotation period of the Galaxy. If the elements 
dispersed into the interstellar medium by supernovae occurring between 
successive compressions are all condensed out to stars at each com- 
pression, the average time these elements are free in space would be 
roughly half the interval between compressions. If, as is more likely, 
just a fraction is condensed at each compression, the average time the 



elements are free in space would be longer, but would presumably still 
be measured in hundreds of millions of years — just about the length of 
time actually determined by nucleocosmochronology. 

Nucleocosmochronologists (if, indeed, there are any astronomers who 
put up with the inconvenience of short gaps for "occupation" on their 
passports and call themselves this) can also estimate the total time since 
the formation of the r-process elements by looking at how much of two 
kinds of uranium exists now. Uranium-235 decays on a time scale of 
710 million years whereas uranium-238 lasts much longer, 4.5 billion 
years. Both are formed from r-process elements in roughly equal 
amounts. When measured in rocks now uranium-238 predominates 
138 times over uranium-235. It can be calculated that approximately 
6.5 billion years must have elapsed since both kinds of uranium were 

The gold you own as a wedding ring, watch or tooth filling is thus the 
seed of some supernova which occurred some six and a half billion years 
ago, which drifted in space for up to half a billion years, which con- 
densed as part of a protosun but was rejected from the body of the 
Sun, and became part of the solar nebula which condensed into the 
planets, including the Earth, and which found itself by chance geological 
processes near enough to the Earth's surface to be extracted by gold 

Metal-rich, metal-poor 

Although the vast majority of stars which we see now con- 
tain the different elements in relative proportions which closely match the 
proportions in the Sun, astronomers have found stars which are deficient 
in the total amount of elements other than the hydrogen and helium 
produced at the origin of the Universe. 

Thus, although, say, the ratio of the amount of iron to nickel is the 
same in most stars, the total amount of iron is highly variable from star 
to star. 

Ignoring chemistry, astronomers have traditionally called all element^ 



In these graphs of the spectra of two stars the black areas {valleys) represent 
colors lost from the star's output of light because of absorption by metal atoms in 
its atmosphere. The bottom spectrum is of HD 8g4gg, a star with a deficiency of 
metals and therefore showing relatively little loss of light due to absorption. A 
heavier blanket of absorption lies across the top spectrum, which is of star HD 
26162 with a high metal content similar to that of the Sun. Olin Eggen ofMt. 
Stromlo Observatory first drew attention to HD 8g4gg, a star formed during the 
initial collapse of our Galaxy, perhaps 8 billion years ago. Mike Bessell estimates 
that it contains only 1% of the concentration of metals in the Sun. 

This diagram is based on spectra taken by Bob Fosbury and Mike Penston 
with the 150-inch Anglo-Australian Telescope. 

other than hydrogen and helium "the metals," and stars with less than 
normal amounts of iron and so on are termed metal-poor. The lines of 
iron and other metals in their spectra are weaker than in other com- 
parable stars signifying that there is a smaller quantity of metals present 
to produce the spectral lines. Conversely metal-rich stars have stronger 
than normal metal lines. 



From the scenario of element creation that has already been sketched, 
the reader can guess that the older stars are metal-poor, having formed 
early in the history of the Galaxy before there had been many supernovae 
to make metals, while the very youngest stars are metal-rich, having 
formed recently from interstellar material which has been enriched by 
the detritus of supernovae throughout the whole of galactic history. The 
Hyades star cluster, that prominent V-shape in the constellation Taurus 
the Bull, is an example of a metal-rich cluster of stars having twice the 
concentration of metals of the Sun and an age of just half a billion years, 
compared with ages in excess of 10 billion years for the oldest metal-poor 

There are, however, no stars found having no metals at all, so the 
amount of metals in the Galaxy appears to have increased steadily from a 
non-zero initial value. In fact, the number of metal-poor stars in general 
is embarrassingly low for simple scenarios. 

Where did the initial metals come from if they were not created in the 
Big Bang ? Where have the metal-poor stars gone ? Or were so few formed 
that very few still exist ? 

One possible answer to these questions is that early in the formation 
of our Galaxy there may have been a wave of star formation which 
preferentially produced large numbers of massive stars. Since massive 
stars quickly turn into supernovae, there would have been a burst of 
metal formation, and the wave of stars would soon have disappeared by 
becoming invisible black holes or faint neutron stars, pulsars whose 
ticking has long since faded to silence. This would imply that in the 
early years of the Galaxy there was a supernova every few days rather 
than every few decades as now. 

With this answer it is difficult to maintain a balance between making 
this first generation of stars massive enough to do a convenient quick 
disappearing act, but not making them so massive that they overproduce 
metals which would still be around now in embarrassing overabundance. 
Among the solutions proposed to cope with this problem is a guess that 
primordial material left over from the formation of the Universe has 



been, and presumably still is, falling into our Galaxy to dilute the metals 
in the interstellar medium. No evidence for the existence of this infalling 
gas is known. Indeed, if there were a large number of supernovae during, 
say, the first two billion years of the existence of the Galaxy, significant 
amounts of gas would be blown out of our Galaxy, rather than falling 
into it. 

You, the supernova remnant 

In this subject a large theoretical superstructure has been 
erected on a narrow keel of observational fact and it cannot be said that 
the so-called Ultimate Model of the chemical history of our Galaxy has 
yet been conceived, let alone detailed. (It is, apparently, called the Ulti- 
mate Model so that when asked what it is, astronomers can reply 

"Urn ") But, according to Beatrice Tinsley in 1974, the scenario 

may proceed in five acts: 

Act I. A Big Bang produces hydrogen and helium, in the early dense 
hot stages at the beginning of the Universe. 

Act II. The Universe cools, becomes less dense and the lumps in it 
start to form galaxies. Actually, the lumps form galaxies more easily in 
the denser early stages than this later one, but nucleocosmochronologists 
turn a blind eye, a deaf ear and a cold shoulder to this difficulty. 

Act III. Galaxies form into spiral or ellipticals depending on how 
they rotate. A portion of the gas in our Galaxy disappears during the 
first two billion years, forming new faint stars of low metal content, and 
the remainder collapses to a flat disk whose metal abundance is by then 
already half that of the solar system because of the first wave of super- 

Act IV. Our Galaxy continues to produce stars at a slowly decreasing 
rate for the next billion years (in the middle of which it produces the 
Sun). The rate of star formation drops slowly because gas is continually 
being locked up into neutron stars, white dwarfs and black holes, but the 
metal content of the gas remaining is continually increasing because each 
supernova enriches the gas with more metals. 



Act V. The Sun having formed, a portion of the metals trapped in the 
rejected part of the solar nebula gives rise to the Earth, biological 
chemistry, evolution and the animal kingdom, including astronomers 
and readers of books. Looked at like this, people are the most interesting 
supernova remnants of all. 

table v Creation of the elements 



Some of the elements 


Hydrogen burning 
Helium burning 

(triple-alpha process) 
Carbon burning 

Origin of Universe 

(Big Bang) 
Most stars 
Red giant stars 

Massive red supergiants 

Hydrogen, helium 

Carbon, oxygen 

Neon, sodium, magne- 
sium, silicon 


/Hot centers of stars ? 
\Supernovae ? 
Evolved stars 

Iron, nickel, chromium, 

Copper, zinc, lead, 



Surfaces or shells of 

Small quantities of 





Gold, platinum, "rare 


(Surfaces of stars ? 
\ Cosmic rays ? 

Lithium, beryllium, 


XII Cosmic rays: 

supernovae and evolution 



•he crackle you hear and feel when you take off a sweater 
of synthetic fiber on a dry day is caused by static electricity. In the act 
of sliding the sweater over your head you rub its surface and fracture 
the atoms there, ionizing some and thus splitting off free electrons from 
the broken atoms. Some of the free electrons can be transferred from 
the sweater to you. You become charged with negative electricity, the 
sweater with corresponding positive electricity. If conditions are right 
your hair stands on end. The spare electrons from the sweater are feeling 
one another's repulsion, pushing their fellows as far away as possible, 
separating your hairs as much as they are able and making them stand 
on end like a porcupine's quills, or Hamlet's hairs when hearing his 
father's ghost's tale of purgatory. 

You may touch a metal light switch and the electrons on you will flee 
into the switch and into the Earth, scattering away from each other as far 
as they can. But suppose you stand still ; suppose you are wearing rubber- 
soled shoes through which the electrons cannot pass. As you watch in 
the mirror you see your hair gradually settle into place as the excess of 
electrons in your body gradually leaves it. In a matter of minutes they 
have dissipated. 

How did the electrons on your body dissipate into the air ? Dry air is 
a very good insulator, like rubber — electrons do not pass easily in air. 



What conducting threads reach through air from you to the Earth, along 
which electrons may pass ? 

Attempts to understand this phenomenon started in 1900. The electron 
had been discovered three years earlier by J. J. Thomson and its role in 
the phenomenon of static electricity was being investigated by Elster 
and Geitel in Germany and C. T. R. Wilson in England. They estab- 
lished that even dry, pure air was not made up solely of complete atoms 
but that some of its atoms had broken into fragments consisting of 
electrons, and what was left, ions. Discharge of static electricity takes 
place by electrons skipping from ion to ion to reach the ground. But 
what causes the ionization of air ? 

The early experimenters attempted to find how the degree of ioniza- 
tion of air varied with atmospheric conditions, geographical location and 
time of day. For this purpose they developed a carefully insulated device 
known as an electroscope which would retain static electricity well 
except for the electrons that it discharged through the air. By watching 
its rate of discharge they hoped to obtain a clue to the amount of ioniza- 
tion of the surrounding air. 

Suspecting that ionization of air was caused by radioactive rocks such 
as uranium-bearing ores, Wulf and Gockel took an electroscope to lakes 
and glaciers which, being of relatively pure water and ice, generated 
little radioactivity. The ionization fell considerably. Some ionization of 
air must therefore be caused by radioactivity. There was however a 
degree of ionization even over the thickest glacier, the deepest freshwater 
lake. Was there some residual radioactivity even in fresh water? Did 
moving the electroscope higher above the glaciers or lakes diminish the 
residual effect ? Curiously it did not. 

If natural radioactivity from rocks was the cause of the ionization, did 
this effect diminish with altitude ? Wulf took his electroscope to the top 
of the Eiffel Tower in Paris (330 meters high). There was some decline 
but not enough. V. F. Hess in Vienna decided to investigate the effect of 
altitude by ascending in a balloon. His first flights to 1000 m showed a 
small reduction of ionization in the air. By 1912 he had made ascents to 


COSMIC rays: supernovae and evolution 

more than 5 km and, to his astonishment, he found that the ionization 
actually began to increase at heights above 2 km. He concluded that 
radiation must be penetrating the atmosphere from above. The obvious 
source of such radiation was the Sun, but this had to be ruled out when 
Hess made an ascent during a partial eclipse. Although the Moon had 
obscured part of the Sun, there was no drop in the ionization of the air. 
Similarly, there was no difference between daytime and nighttime levels. 
Nor could the 23 hour 56 minute period of the Earth's spin beneath the 
stars be picked out — so whatever space radiation was ionizing the air 
could not come from any one object. 

Explaining cosmic rays 

These mysterious emanations were first termed cosmic 
radiation by the great American physicist R. Millikan, who began 
research in the subject after World War I. He also realized that cosmic 
rays were bringing an energy comparable to that of starlight to the 

Gradually the complex properties of the cosmic rays became under- 
stood. The modern picture is that the cosmic rays incident on the Earth 
from space are atoms, stripped of most or all of their electrons and 
therefore are bare nuclei, moving at speeds close to that of light. 

Cosmic rays are influenced by the magnetic forces in the Galaxy and 
the solar system. The lowest energy cosmic rays are deflected by 
magnetized clouds ejected from the Sun during periods of solar activity. 
The Sun's activity varies with an 1 1 year cycle. During its most active 
phase there are many large sunspots on its surface and many solar 
"storms" which eject the clouds into the space between the planets. 
These clouds cocoon the Earth and shield it from the influence of 
interstellar cosmic rays, though the Sun itself may contribute some 
extra cosmic rays to the flux striking Earth. The magnetic field of the 
Earth itself also plays a part in deflecting cosmic rays. 

Those cosmic rays which do reach Earth's upper atmosphere collide 



with air atoms to make showers of electrons and other particles of matter 
which cascade down to the Earth's surface. These secondary particles 
are the ones which ionize air near sea level and cause the gradual dis- 
charge of static electricity from electroscope or tousled hair. 

Much of the early study was of the secondary cosmic rays; only as 
balloons and rockets reached far above Earth did the properties of the 
primary cosmic rays become clearer. Present-day studies have used 
cosmic ray detectors in space, both on artificial satellites and on the 
Earth's natural satellite, the Moon. The latter experiments were set 
up on the Moon's surface during the Apollo 16 lunar flight and brought 
back to Earth afterwards. 

The abundances of the different nuclei found among the cosmic rays 
follow closely the abundances of the nuclei determined from the cosmo- 
chemistry of meteorites and the like as discussed in the last chapter — 
hydrogen is the most abundant, and the 90-odd remaining heavier 
elements are less and less common as they increase in complexity. There 
are three major exceptions — the elements just heavier than helium, 
namely lithium, beryllium and boron, are a million times more abundant 
in the cosmic rays than in other matter in the Universe. 

We saw in the last chapter that the creation of elements in the Big 
Bang could not get to lithium. Moreover, it turns Out that nuclei of all 
three of these elements are readily destroyed at temperatures of about a 
million degrees — relatively low compared with those found in stars. 
Far from lithium, beryllium and boron being created in stars, therefore, 
any interstellar material condensing into a star and then being returned 
to space in a supernova or whatever has been purged of these elements, 
unless they have been made on the relatively cooler stellar surface or in 
space itself. If these three elements can be made neither in the Big Bang 
nor in stars, where can their origin lie ? The B 2 FH paper of the Bur- 
bidges, Fowler and Hoyle named the process which created these 
elements the x-process because it was unknown. 

The ^-process may in fact be the travel of cosmic rays through the 
interstellar material. The cosmic rays contain carbon, nitrogen and 


COSMIC rays: supernovae and evolution 

oxygen nuclei in relatively high abundance. As such nuclei collide with 
hydrogen and helium in interstellar space, the individual protons and 
neutrons in the nuclei which collide are rearranged into several fragments 
of different sizes including lithium, beryllium and boron nuclei. (This 
process is also called spallation.) The cosmic rays might therefore be 
expected to contain an excess of spallation products. They contain not 
only the great excess of lithium, beryllium and boron, but also other 
spallation elements such as chlorine and manganese. 

From the amount of these elements in cosmic rays, cosmic ray 
physicists can establish through how much interstellar matter the average 
cosmic ray has traveled so as to produce the observed quantity of 
spallation products. The average cosmic ray has been traveling in the 
Galaxy for some 10 to ioo million years. After this time it leaks from the 
Galaxy into intergalactic space. Though a long time on a human time 
scale, 10 to ioo million years is short when compared to the age of the 
Universe (measured in units of 10 billion years) and the lifetime of most 
stars. If it is to the stars that we must look for the creation of cosmic 
rays, it is to the short lived, massive ones that we must direct our atten- 
tion, and these are the ones which turn into supernovae. In fact Baade 
and Zwicky calculated that supernovae occur often enough and have 
enough energy to account for the supply of all the cosmic rays in the 

The way in which they do so is not clear. Some believe that cosmic 
rays are injected into space during the supernova explosion itself. 
Others say that it is not the supernova which supplies cosmic rays to the 
Galaxy but the rotating neutron star or pulsar, which the supernova 
forms. This sweeps up interstellar material and flings it into the Galaxy 
at great speed. 

Perhaps a mixture of these processes is responsible for pumping 
cosmic rays into space to stream about the Galaxy, and, before they 
escape from it, possibly to collide with Earth. 


The influence of cosmic rays 

Do the cosmic rays have more profound effects on the 
Earth than influencing the discharge of static electricity? In fact it 
appears that they do. Because of supernovae, archaeologists have a time 
scale for dating in prehistory. 

The spallation reactions occurring as the cosmic rays collide with 
Earth's atmosphere produce neutrons, most of which are eventually 
absorbed by the predominant gas in the atmosphere, nitrogen. In this 
way the nitrogen turns into a carbon atom, of a kind called carbon-14 
having an excess of two neutrons over the more usual carbon- 12. Because 
of this, carbon-14 is unstable and decays to nitrogen-14, half of it 
changing every 5568 years. Though the carbon-14 is produced high in 
the atmosphere, at an altitude of 10 to 15 km, it combines with oxygen 
to make carbon dioxide, and diffuses rapidly to Earth's surface. 

Carbon dioxide is a gas which is abundant in the atmosphere and 
which is "breathed in" by plants to take part in their cell building 
processes. In this way, therefore a small proportion of the carbon atoms 
which the plants accumulate during their lifetimes are radioactive 
carbon-14 atoms. Because plants are eaten by animals, all living organic 
material contains this proportion of radioactive carbon. When the plant 
or animal dies, the assimilation of carbon stops and the carbon-14 begins 
its long, steady decay to ordinary carbon. If the organism has recently 
died, its cells will contain their full abundance of carbon-14; if it has 
been dead a long time, the proportion of carbon-14 will be small. The 
ratio of ordinary carbon to carbon-14 is therefore a clock which runs 
down at a known rate. 

In the cases where parts of the dead organic matter remain pre- 
served — as timbers in a house, or bones in a midden, for example — this 
clock can be used to measure with some accuracy the year in which the 
tree was felled or the animal died. This technique, known as radiocarbon 
dating, can be used to estimate the age of material up to 20,000 years old, 
and is therefore of inestimable value in archaeology. 



Now this method relies on the level of cosmic radiation having been 
constant for many thousands of years. Suppose there is a sudden signifi- 
cant burst of cosmic radiation from space, producing more carbon- 14 
than normal. When this carbon- 14 is assimilated by living organisms, 
their radioactivity level is suddenly much higher. After death, the 
organic material still exhibits a proportionally higher radioactivity level. 
It would seem to the archeologist that the clock was still tightly wound, 
and that the organism was more recently dead than was the case. 

There is good evidence that the production of carbon- 14 has indeed 
been variable in the past. Evidence for the variation comes from using 
two different methods to estimate the age of wood from the incredibly 
long-lived California bristlecone pine. 

One method is radiocarbon dating, while the other is called dendro- 
chronology — counting tree rings. Each year, a tree adds a new ring to 
its circumference. Counting these rings offers a potentially accurate 
way of determining the age of a tree whose date of death or felling is 
known, and bristlecone pines as old as 4600 years have been found. By 
overlapping ring sequences from many bristlecone pines, C. W. Ferguson 
has established the age of wood which formed 8000 years ago. Carbon- 14 
dating of wood as old as that, however, seems to suggest that it is some 
900 years younger. Only over the last 2500 years do the two dating 
methods agree on average. Evidently the rate of production of carbon- 14 
was once higher than now. 

Was this the result of a sudden burst of supernova activity increasing 
the number of cosmic rays in the Galaxy? The Czech geophysicist 
V. Bucha offers another explanation : that it was due to changes in the 
Earth's magnetism and its corresponding ability to shield the atmosphere 
from cosmic rays. 

There is no real evidence for a significant change in the cosmic ray 
flux in historical times. The record of "fossil" cosmic rays has been 
extended back to 10 million years by a count of the number of tracks 
left by cosmic rays as they struck meteorites orbiting the solar system, 
showing no major changes in that time interval. Before then, the cosmic 



ray intensity in the Galaxy could have been different from now, and 
presumably, was much larger very long ago during the first wave of 
formation of massive stars in the Galaxy and the consequent high rate of 
supernovae. But we have no direct evidence about the cosmic ray rate at 
this time. 

Did a supernova kill the dinosaurs ? 

In 1957 Iosif Shklovsky considered what would occur if a 
supernova exploded near to the Sun, within say ten parsecs. The super- 
nova would shine at magnitude — 20 for a matter of months (not as 
bright as the Sun but far brighter than the Full Moon). After a few 
thousand years, the gaseous envelope of the supernova, ejected from it, 
would pass across the solar system. If the supernova were like the Crab 
Nebula, the synchrotron radiation trapped within the filaments would 
be seen as bright as the Milky Way, filling half the night sky. There 
would be no dynamical changes in the solar system — the amount of mass 
striking the planets would be too small to deflect them from their orbits 
at all. However, the solar system would be within the supernova shell 
for some 10,000 years, and the density of cosmic rays would be 10 to 100 
times its current value. There would be much more cosmic radiation 
striking the Earth's atmosphere and much more radiation at sea level 
caused by the secondary cosmic rays. 

Shklovsky has pointed out that such an increased level of background 
radioactivity might have played a part in the evolution of life on Earth. 
Evolution proceeds as a result of chance mutations in individual bio- 
logical organisms. Some mutations better fit the organism to survive and, 
during its longer lifetime, it has a better chance to perpetuate the 
mutation to future generations. Most mutations, however, are unfavor- 
able, particularly when the change to the genetic material has been gross. 
It might be expected that a large increase in radioactivity would produce 
many gross changes in genes and thus tend to cause organisms to die and 
become extinct. 


COSMIC rays: supernovae and evolution 

Two thirds of the mean radioactivity on Earth is caused by terrestrial 
factors, mostly natural radioactivity in rocks at Earth's surface. One 
third is caused by cosmic rays. If the cosmic ray intensity increased 
ioo-fold, the background radioactivity would be some 30 times increased. 
Shklovsky speculates that the dying out of the giant reptiles at the end 
of the Cretaceous period was the consequence of such an increase, 
caused by the sudden bathing of the Earth in the ejecta of a close 
supernova unwitnessed save by the uncomprehending eyes of doomed 
prehistoric dinosaurs. 

This speculation about the reason why dinosaurs are extinct is one of 
a class termed "cataclysmic" by paleontologists — the class includes all 
global catastrophes which might exterminate a widespread order of 
animals. But there is no evidence in the geological record for the 
extermination of other orders at the same time. Dinosaurs did not live 
alone and their contemporaries survived them. Paleontologists generally 
therefore accept no cataclysmic reason for the extinction of the dinosaurs, 
and Shklovsky's attractive notion is not widely accepted. 

In an interesting development of the theory that cosmic rays cause 
mutations, George Michanowsky proposed that cosmic rays from the 
Vela supernova, occurring about 8000 bc, triggered off a new awareness 
in men's minds and precipitated the technological era which we are still 
witnessing. If this is right the Vela supernova would be like the intelli- 
gence-donating obelisk in Arthur C. Clarke's book and movie, 2001 — A 
Space Odyssey, and might have had a physical effect on mankind com- 
parable to the mental liberation sparked by Tycho's supernova in 1572. 

This is just speculation. Nonetheless it must be true that cosmic rays, 
by working on the genes of all species, inevitably cause mutations. Since, 
in fact, the natural radioactivity in rocks is a consequence of the forma- 
tion of radioactive heavy elements in supernova explosions, and since 
the cosmic rays are generated in supernovae explosions, it could be said 
that, having played a crucial part in producing and distributing the 
chemical elements which make life possible at all, the supernovae are 
responsible for life's further evolution. 


XIII Supernovae in binary stars 


Tily about 15 per cent of all stars are in what we might 
call the "lone star state" — that is, they are single, like the Sun. Nearly 
half (46 per cent) have partners, with one star orbiting the other; the 
remainder (39 per cent) occur in multiple star systems. 

It must therefore be common for a supernova to occur in a multiple 
star system. 

The star which explodes first in a double or binary star system is the 
larger of the two since more massive stars expend their nuclear energy 
at a faster rate than smaller ones, and pass more rapidly to the late stages 
of stellar evolution. 

What happens when the more massive star of a pair explodes as a 
supernova? Certainly the supernova explosion has some effect. In the 
first place, the smaller companion star receives the impact of the explo- 
sion and recoils like a catcher's mitt grasping a fast-thrown ball. If the 
companion star was describing a circular orbit about the exploding star, 
the orbit is now eccentric or flattened to some degree. The companion 
may also be blasted by stellar shrapnel, and material from the exploding 
star may significantly increase the mass of its companion. Even if it was 
heading to a quiet end as a white dwarf before the supernova explosion, 
it may now itself be too large for this and be destined to be a supernova 
too. In this sense, at least, supernovae can cause supernovae. 



Not all the exploding material accretes onto the companion star. 
Perhaps most is ejected into space. It is even possible that the blast from 
a supernova is so severe that it strips the outer layers from its com- 
panion, decreasing its mass and lengthening its life. All in all, more than 
half the total mass of the double star system can be blown off into space. 
In this case the pair is split asunder, each star flying away like a stone 
flung from a slingshot. As a numerical example, consider a 30 solar mass 
star and a 4 solar mass star orbiting each other, making a total of 34 solar 
masses. The large star explodes leaving a 2 solar mass neutron star 
behind and ejecting 28 solar masses. Eight solar masses of this may fall 
onto the smaller companion, now making it 12 solar masses, while 20 
solar masses of material is ejected into space. Because this is more than 
half the original mass of the double star, the neutron star and the 12 solar 
mass star are flung in opposite directions, speeding away from each 
other. Where there was once a double star, now there is simply an empty 
space at the center of an expanding shell of fragments, with two stars 
rushing away. 

Where might we find such events ? Stars of 30 solar masses are profli- 
gate with their nuclear energy and have lifetimes measured in only a few 
millions of years, compared with the 10 billion-year lifetime of the 
Galaxy. They are thus young objects and must have been recently 
formed from the interstellar gas in the Galaxy. This gas only occurs in 
the plane of the Galaxy in spiral arms. Thus bright, massive stars are 
found near the galactic plane. 

If one star of a pair goes supernova and disrupts the double star, the 
two stars are quite likely to be flung out of the galactic plane altogether. 
This ties in well with the observations of radio pulsars, which are indeed 
found high above the galactic plane. The nine whose proper motions 
have been measured have an average speed in excess of 250 km/s, com- 
pared with a typical average speed of 15 km/s for stars just formed from 
the interstellar material. The evidence thus suggests that radio pulsars 
are formed from a binary star system during a supernova explosion 
which disrupts the binary star. 



This is confirmed by the fact that, in contrast to the average popula- 
tion of stars where a majority are in multiple star systems, only one of the 
149 pulsars known is in a double star. Even this lone example of a pulsar 
in a binary star is describing a very eccentric and elongated orbit about 
its stellar companion, suggesting that it only just failed to be severed 
from the double star system. 

What of the former companions of the pulsars ? Among the normal 
slow-moving population of bright, massive stars in our Galaxy there are 
some so-called "runaway" stars, with high speeds often exceeding 
100 km/s. Three of these are Mu Columbae, AE Aurigae and 53 Arietis. 
From their speeds and directions, astronomers calculate that each left 
the region of the constellation of Orion some three million years ago, 
though they are now in the widely separated constellations of Columba, 
Auriga and Aries. Orion contains a large number of massive stars, and 
it was postulated by Adriaan Blaauw in 1961, following a suggestion by 
Fritz Zwicky, that these three stars were in a quadruple star system, the 
fourth member of which exploded as a supernova flinging the three 
runaway stars far from their birthplace. 

Many astronomers believe this to be the general explanation for the 
massive runaway stars. It is also true that the runaway stars themselves 
will eventually undergo a supernova explosion, turning them into pulsars 
like their former companions, and they will then continue in their high 
speed flight beyond the plane of the Galaxy. 

Some people have speculated that the Crab Nebula supernova was 
originally a runaway star formed by the earlier disruption of a binary 
star in the supernova explosion that produced another pulsar near to 
the Crab Nebula, namely np 0527, which has the peculiarity of having 
the slowest pulse rate of any pulsar. The hypothesis was that both the 
Crab Nebula pulsar and np 0527 were once members of a binary star 
system in the association of massive stars called I Geminorum. However, 
the problem with this attractive idea is that the space velocity of np 0527 
is too slow for it to have traveled from this association to where it now 
lies, within the short lifetime of a pulsar. The hypothesis is not well 


Runaway stars AE Aurigae, 53 Arietis and Mu Columbae were thrown like 
slingshots from the constellation Orion three million years ago when a fourth star 
of a quadruple star system disintegrated in a supernova explosion. Each has 
traveled right across an intervening constellation to reach the present-day 
positions 30 to 40 degrees from the origin. 



substantiated. However the Crab Nebula pulsar itself does have a 
considerable proper motion away from the constellation Gemini and it 
may be that the supernova of 1054 occurred on a runaway star ejected 
about four million years ago from the I Geminorum association of 
massive stars. 

It is not inconceivable that the momentum of the runaway stars and 
pulsars will in some cases be sufficient to carry them right out of the 
Galaxy. If the speed of the pulsars near the Sun exceeds some 290 km/s 
they have more than the velocity of escape from the Galaxy and will 
escape from it into intergalactic space. Pulsar number 0283 + 36 probably 
has such a speed and may be doomed to become an intergalactic tramp. 

Presumably not every binary star system in which a supernova explo- 
sion takes place is disrupted. Sometimes the amount of material ejected 
will be less than half the total mass of the double star system so the laws 
of physics dictate that the pair will not break apart. 

Perhaps the explosion itself will cause the pulsar formed to recoil in 
such a way that it remains in orbit around its companion. Why is it then 
that binary pulsars are so rare ? The answer is presumably that the huge 
envelope of ionized gas from the ordinary star surrounds the pair and 
traps any detectable radio emission from the pulsar. Such a cloud of gas 
is blown off by our own Sun, and is known as the solar wind. However 
the pulsar does ultimately become detectable — as an X-ray pulsar. 

Why a pulsar shines X-rays 

In the course of time the hydrogen fuel in the center of the 
companion of an unseen pulsar in a binary star system gives out and the 
star begins to expand into a red giant (as outlined in Chapter IX). It 
may be that the star and pulsar are sufficiently close that before the star's 
growth to a red giant is complete its atmosphere begins to leak onto the 

It is easy to see that at a certain point between two stars their gravita- 
tional forces exactly cancel each other, so that an atom placed at this 



point might find it difficult to decide to which star it should fall. Jules 
Verne wrote a story in which the crew of a giant shell fired into space 
suddenly fell from the floor to the ceiling of their capsule as they passed 
the equivalent point between Earth and Moon. (He did not understand 
the concept of free fall, and that astronauts are weightless except when 
their space capsule is being propelled.) The real situation in a binary 
star is complicated by the presence of centrifugal forces caused by the 
revolution of the two stars in orbit around each other, but nonetheless 
the gravitational field of each star has its own zone of influence within 
which all material belongs to that star. These areas are teardrop shaped 
with the points touching, and are called Roche lobes. When one star fills 
its Roche lobe its atmosphere may protrude beyond its lobe into the 
adjacent one and fall towards the other star. So gas is transferred from 
the putative red giant to the pulsar. 

Now, the pulsar is small : it is a neutron star, some 20 km in diameter. 
When the gas from its companion falls upon the neutron star, the gas is 
compressed by the intense gravitational force. Just as compressing the 
air in a bicycle pump heats it, so the infalling gas is heated, to a tem- 
perature which may be tens of millions of degrees. The hotter a body 
the more energetic (shorter wavelength) the radiation it emits. This gas 
is so hot that it shines not by emitting infrared radiation or light but by 
emitting X-rays. The neutron star becomes detectable as an X-ray star. 

The power available from the gravitational field of a neutron star is 
enormous. A marshmallow falling onto the surface of a neutron star 
would explode with the energy release of a World War II atomic bomb. 

Many X-ray binary stars are now known from observations by X-ray 
survey telescopes on board artificial satellites, particularly Uhuru, 
launched in 1970, and the Copernicus satellite launched in 1972. 

How do astronomers know that a particular X-ray source is part of a 
binary star ? Take as an example the strongest X-ray source in Hercules, 
called Her x-i. This source switches off for 6^ hours every 1.7 days. 
Similar behavior is found among ordinary stars, too, the explanation 
being that they are double stars of the type called eclipsing binaries. Quite 



simply, one star hides the other for a while during its orbit because the 
plane of their orbits is roughly in our line of sight. 

In the case of Her x-i, the companion was known to be a variable, 
called HZ Herculis, before the X-rays were detected coming from the 
source. It is variable because the X-rays shining from the neutron star 
heat one side, this side turning towards and away from Earth with the 
orbital period of the neutron star. Her x-i itself is a pulsar, pulsing 
X-rays with a period of nearly one and a quarter seconds. As the pulsar 
orbits HZ Herculis, the Doppler shift of its pulsing frequency can be 
clearly detected. In fact, the Doppler shift of the companion star can 
also be measured, since some of the pulsed X-rays are intercepted by the 
companion star where they heat up its surface and are re-emitted as 
pulsing visible light. 

A sustained and difficult monitoring of this weak pulsation has been 
made by Berkeley astronomers Jerry Nelson and John Middleditch so 
that they have been able to measure the Doppler shift of the light 
pulses emitted from the companion star and determine its orbit about 
the neutron star. 

From these measurements, astronomers can tell at exactly how many 
kilometers a second each star is traveling. Combining this with the 
orbital period of 1.7 days and the fact that the orbits are in our line of 
sight, all details of the orbits are known. The importance of this is that 
the masses of the two objects can be worked out, using the same orbital 
laws that Kepler deduced from observations of the solar system. 

At last astronomers have been able to get to grips with observational 
facts about a neutron star. Theoreticians had calculated that no neutron 
star could have a mass greater than twice that of the Sun. In vindication 
of their work, the mass of Her x-i turned out to be 1.3 solar masses. 

Unlike radio pulsars which are all slowing down, the two best studied 
X-ray pulsars, Her x-i and Cen x-3, are speeding up. This is probably 
because X-ray pulsars are members of binary systems by their very 
nature. Impact of the infalling matter onto the surface of the neutron 
star gives an impulse to the star, speeding it up. 


A nova flares 

Astronomers now believe that processes in which close 
binary stars feed upon one another are quite common and are responsible 
for another type of stellar cataclysm — novae. In recent years plausible 
theories have been put forward which link novae with events in close 
binary systems consisting of a red giant and white dwarf. 

We have seen that double stars are common, and have shown how 
material can transfer from one star to another when a red giant star 
expands outside its Roche lobe. When the material falls onto a pulsar it 
heats up and emits X-rays, creating a pulsating X-ray source. But when 
material falls onto the already hot surface of a white dwarf star, the 
material may build up to form a shell of hydrogen within which stellar 
nuclear reactions may take place. The surface then explodes, and a nova 

The total mass thrown off the white dwarf is typically only a few 
millionths of the original mass of the binary star system. For all its 
drama, and unlike a supernova explosion, the nova phenomenon does 
not penetrate deeply into the white dwarf. Nonetheless the mass thrown 
off is sometimes detectable as it causes a change in the period of the 
binary star. The nova in the constellation Hercules in the year 1934 was 
known as an eclipsing binary star with a period of 0.1932084 day before 
its outburst. Afterwards its period had lengthened to 0.1936206 day. 

Gas continues to flow from the red giant onto the white dwarf after a 
nova explosion and there can be a further buildup of hydrogen and 
repeated nova explosions. Stars in which more than one nova explosion 
has occurred are called recurrent novae; it may be, given a long enough 
time, that most novae are recurrent. A recurrent nova in the southern 
constellation Pyxis holds the record for the number of outbursts, with a 
total of 4 in the last 80 years. 

Astronomers are pretty sure that they are on the right lines in talking 
about close binary systems as an explanation for both pulsating X-ray 
sources and optical novae. But almost monthly, X-ray satellites such as 



Ariel 5 are picking up new kinds of flaring sources which astronomers 
term transient stars, and where these fit into the picture is not clear. 

There are the bursters, which flare up within seconds and fade away 
quickly, so the whole event is over within a couple of minutes. Some 
sources flare and die regularly, with a large flare being followed by a long 
quiet period, while a small flare is followed by a shorter quiet period. 
It is very likely that such behavior is caused by a buildup of material in 
the flow from one star to another and its sudden heating. 

Then there are the X-ray novae, which can dominate the X-ray sky 
for a few weeks at a time. Only two of these have, at the time of writing, 
been linked with optical novae. Probably what is seen is the side of the 
companion star facing the neutron star, heated by the outburst. As the 
X-ray nova dies away so the companion star cools and its burst of light 
fades too — the fading shadow of a brief candle. 

table vi X-ray binary stars 


Star name 




Mass Mass of Nature of 





of star com- companion 




(units of Sun's mass) 

Her x-i 

HZ Her 




2 1.3 neutron star 

Cen x-3 





18 1 neutron star 

SMC x-i 





28 about 2 neutron star 

Cyg x-i 

HDE 226868 




15 10? black hole? 


XIV Black holes from supernovae 

The bigger they come, the harder they fall. 
R. Fitzsimmons, 
champion boxer 


lot all supernovae produce neutron stars. Searches 
at the centers of supernova remnants formed by recent supernovae such 
as Tycho's, Kepler's and the Cas A supernovae remnants have revealed 
no pulsars. Perhaps the beams of these pulsars never point to Earth in 
their rotations. But there are too many supernova remnants devoid of 
pulsars for this to be the likely explanation for all. Perhaps supernovae 
produce other, stranger stars which we cannot see. 

How to escape 

A stone tossed in the air loses speed and momentarily halts 
at its highest point before plummeting back to Earth again. When 
hurled with greater force, the stone rises higher but is still drawn back 
by the force of gravity. There is clearly a relation between the initial 
speed given to the stone and the height to which it rises: scientifically 
speaking, that momentary halt at maximum height is where the stone's 
potential energy due to the force of gravity exactly equals the kinetic 
energy it was given as it was thrown. The question arises : what gravita- 
tional energy does the stone have at higher points above the Earth and 
what speed is it necessary to give the stone to throw it arbitrarily far 
above the Earth, to "infinity"? Is it possible for the stone to be hurled 



with sufficient speed for it to leave the Earth completely and escape its 
gravitational pull? It is, and this speed is called the velocity of escape. 
The velocity of escape from the Earth's surface is 25,000 miles an hour: 
an object thrown with this speed from Earth does not fall back. 

How does the velocity of escape vary from place to place — from planet 
to planet, star to star? The larger an astronomical body, the less the 
force of gravity at its surface; but the more massive it is, the greater its 
gravitational force. Put in mathematical terms, the velocity of escape is 
proportional to the square root of the mass of the body divided by its 
radius. Jupiter has 318 times the mass of the Earth and a radius 1 1 times 

150,000 STAR 

Comparative sizes and 
escape velocity of the Sun, 
Earth and some stars. 
Escaping from Earth needs 
an 11 km Is impulse and is 
easiest because the Earth's 
mass is small. A red super- 
giant is the easiest star to 
escape from as its size is 
large. Escaping from other 
stars becomes more difficult 
as the stars become more 
compact. Escaping from the 
most compact objects known, 
black holes, is impossible, 
requiring a speed greater 
than that of light. 



larger than the Earth's: the escape velocity from Jupiter is therefore the 
square root of 318/n, which is 5.4 times that from Earth. 

The escape velocity from the Sun is more than a million miles an 
hour : though larger than the Earth it is much more massive and so has 
a larger velocity of escape. Curiously, although other dwarf stars like 
the Sun range in mass from a few hundredths of the Sun's mass to 60 
times its mass, the range of sizes almost exactly compensates for this, 
with the result that the escape velocity from all other dwarf stars is not 
much different from that of the Sun. 

Only when stars have finished the phase in which they burn hydrogen 
and expand to become red giants and supergiants does their escape 
velocity alter much from the solar value. In a red supergiant the escape 
velocity can be less than one sixth the solar value. This is the reason why 
matter can relatively easily escape from stars at the red giant and super- 
giant stage so that some massive stars can bring their mass under the 
Chandrasekhar limit and become white dwarfs in their subsequent 

As you might expect, white dwarfs have high escape velocities. Even 
though* their masses are similar to that of the Sun, they are denser and 
more compact, with radiuses typically one hundredth the Sun's. From 
the square-root law, therefore, the escape velocity is typically ten times 
as high as the Sun's. If matter is to escape from a white dwarf, such as in 
a nova explosion, when the white dwarf's outer envelope is thrown off, 
the explosion must be energetic enough to give the envelope an escape 
velocity of thousands of kilometers per second. 

The concept of escape velocity is useful in describing not only the 
energy required to blast matter off the surface of a star, but also the loss 
of energy by radiation as it is emitted from the star. Radiation too loses 
energy as it travels against the force of gravity. This fact is not obvious 
from ordinary experience but is a feature of Einstein's General Theory 
of Relativity. Although very small in the Earth's gravitational field, the 
loss of energy has nonetheless been measured in gamma rays traveling 
up a mine shaft. 



The lower the energy of radiation, the longer its wavelength. As light 
loses energy climbing out of a gravitational field, therefore, its color 
shifts towards the red end of the spectrum. This phenomenon is termed 
the gravitational redshift. The fraction of energy lost by radiation as it 
leaves the surface of a star is the square of the velocity of escape measured 
as a fraction of the speed of light. The gravitational redshift is not 
measurable in most stars, as we can see by taking the Sun as an example. 
Since the velocity of escape from the Sun is about 600 km/s and the 
speed of light is 300,000 km/s, the fraction of energy lost by light as it 
leaves the Sun's surface is only (600/3 oo,ooo) 2 or one part in a quarter 

But although the Sun's gravitational redshift is barely detectable, soon 
after white dwarfs had been discovered Arthur Stanley Eddington in 
1924 pointed out they they had a large escape velocity and that measur- 
able redshifts could be expected in their spectra. The escape velocity 
from a white dwarf is typically 3000 km/s (Table VIII), so the redshift is 
(3000/300,000)% or one part in 10,000. Thus H-alpha light emitted from 
hydrogen at a dense white dwarf's surface with wavelength 6563 
angstrom units might be seen by an observer on Earth with wavelength 
almost 6564 angstroms, a small but measurable increase. There was, 
however, a confusing detail : there is no way of distinguishing a gravita- 
tional redshift from the more familiar Doppler redshift caused by possi- 
ble motion of the white dwarf away from Earth, at a speed of a few tens 
of kilometers per second or so, unless the speed could be accurately 
known, and allowed for. 

The only way round this problem was by finding white dwarfs which 
were part of binary systems, sharing a common motion with other 
ordinary stars. W. S. Adams measured the shift in the spectrum of the 
white dwarf Sirius B, which is in orbit around the bright star Sirius 
itself. The speed of Sirius is 8 km/s towards the Sun, the speed of motion 
of Sirius B around Sirius could be accurately calculated from knowledge 
of its orbit, and the redshift still unaccounted for was 21 km/s compared 
with the value of 20 km/s calculated by Eddington. The redshift of 



40 Eridani B, a white dwarf in orbit around the star 40 Eridani, has been 
similarly measured by D. M. Popper, with good agreement with the 
theoretical value. 

Measuring the gravitational redshift from the surface of a neutron 
star such as the Crab Nebula pulsar would be very interesting, since the 
wavelength of light emitted from its surface would be much changed 
from its original value. Formerly invisible ultraviolet light would be 
redshifted into the visible part of the spectrum, and the shift would give 
astronomers a way to estimate the mass and radius of a pulsar. In fact 
there are no atoms at the surface of a neutron star to emit light of a 
distinct wavelength: the redshift cannot be measured, although it 
would be very large. 

Possibly enormous gravitational redshifts may be responsible for the 
behavior of high redshift quasars, mysterious starlike objects lying 
among other galaxies. Some, perhaps most, astronomers argue however 
that the redshift of quasars is caused by their motion in the Universe as 
they participate in the explosion of the Big Bang and recede from us, 
and that the high values of the redshifts of quasars are a consequence 
of their great distance. 

There is a limit to how large a gravitational redshift can be. The 
radiation cannot lose more energy than it possesses in climbing the 
gravitational field of a star. The ultimate redshift occurs when the frac- 
tion of energy lost by radiation is 100 per cent, which occurs when the 
escape velocity at the star's surface is equal to the speed of light. It is 
possible, then, to conceive of a star with such a powerful gravitational 
field that radiation cannot leave the star. Nor can matter leave the star 
since to do so it would have to travel at the velocity of light, and accord- 
ing to Einstein's Theory of Relativity nothing material can travel at the 
speed of light or faster. 

Nothing at all could ever leave such a star because its gravity would 
be simply too strong — it would be black because no light could leave it, 
and it would be a hole because anything dropped in could not get out. 
Hence the name of such a star — the black hole. 


How to make a black hole 

How would such a star be formed ? Take an ordinary star 
(in what is called a "thought experiment," one which is impossible 
actually to carry out except in imagination) and compress it in a vise in 
an attempt to make it smaller and thus increase its escape velocity and 
its gravitational redshift. The star will resist this attempt by increasing 
its internal pressure — reducing the star's size causes its atoms to pound 
faster and more often on the jaws of the vise to attempt to pry it apart. 

The star will remain in equilibrium in a kind of war between nar- 
cissistic gravitational self-attraction and the incest-taboo of its repelling 
internal pressure forces. 

But suppose that the gravitational vise clamped tighter and tighter, 
squeezing the interior of the star smaller and smaller. A neutron star 
might be the result. It too would resist further compression but less 
enthusiastically than ordinary stars: the repelling pressure mechanism 
is not so easily able to respond to any increased self-attraction. In fact 
the more massive neutron stars are less able to increase their internal 
pressure in response to further gravitational contraction. Beyond a 
certain critical mass in fact, they cannot respond at all: their internal 
pressure is at a maximum and cannot be increased. The gravitational 
self-attraction of such a star is always larger than the repelling pressure 
forces. The star cannot support itself and simply shrinks smaller and 
smaller in a continuing collapse. As it does so the velocity of escape at its 
surface increases until it reaches and passes the speed of light, at which 
stage whatever is inside the surface is shut off from the rest of the 
Universe forever. The star has become a black hole. 

Since, in a sense, a black hole can be formed as a kind of extreme 
neutron star, it is natural to look to the same process which forms neu- 
tron stars to form black holes. Black holes appear to be formed in super- 
nova explosions in the case where the mass of the stellar core which 
begins the collapse exceeds the critical mass of neutron stars, a few solar 


Are black holes the missing mass ? 

Is there evidence that black holes exist ? There is certainly 
evidence for the existence of invisible matter in our Galaxy. Its effect 
can be seen in the motions of stars. As it orbits the Galaxy, a star is 
subject to the gravitational pull of all the others in the Galaxy. The stars 
which we see away from the Milky Way are above or below the Sun, 
which lies close to the central plane of the Galaxy wherein lie most 
stars. Stars above or below the galactic plane are pulled back towards it 
and the force which pulls them can be estimated from the speeds of 
stars at different heights above the plane. The higher stars generally 
move more slowly, just like stones flung from the surface of the Earth. 
If the force pulling these stars back is known, we can calculate the 
amount of mass in the galactic plane required to produce such a force. 
The answer is that in every cube of space of 10 light years on the side 
near the galactic plane in the vicinity of the Sun there is on average 
4.5 solar masses of material. About 1.8 solar masses of that can be 
accounted for as visible stars. A further 0.9 solar masses can be detected 
as interstellar gas in the form of hydrogen atoms. Approximately 2.7 
solar masses is invisible. Part of this mass is certainly hydrogen molecules 
which emit no identifiable radiation received on Earth; part is un- 
doubtedly neutron stars which have stopped being pulsars and part is 
former white dwarfs which have cooled to invisible black dwarfs. There 
has been speculation that some fraction at least of this so called " missing 
mass" in the Galaxy is in the form of black holes. Potentially, then, there 
may be large numbers of black holes waiting to be found. 

But, "of all objects that one can conceive to be traveling through 
empty space," wrote R. Ruffini and J. A. Wheeler in 1971, "few offer 
poorer prospects of detection than a solitary black hole ..." By its very 
nature the black hole cloaks itself with invisibility. No evidence of it 
existence beyond its surface can ever be seen, no action on its surface can 
send a message to proclaim its occurrence. The message carrier— radio 
pulse, light flash, cosmic ray or what you will — cannot escape the black 



hole's pull of gravity, which is why the surface of a black hole is called 
the event horizon. Occurrences within it are never seen. 

If we cannot see a black hole itself can we see its surroundings ? What 
happens when something encounters a black hole ? What would happen 
if a black hole were to draw material into itself by gravitational force ? 

How the invisible shines 

In an effort to describe how a black hole could be found 
Iosif Shklovsky in 1967 considered what would happen if a star were to 
orbit a black hole. The black hole and its companion would circle each 
other with little effect on each other at first, beyond their mutual 
orbiting. A distant astronomer might wonder why the radial velocity of 
the companion star changed periodically and deduce that the star was 
orbiting another which he could not see. (Such stars are known as 
single-line spectroscopic binaries and it is possible that some of their 
invisible companions are black holes, though no doubt the vast majority 
of invisible companions are simply fainter stars outshone by the star 
which can be seen.) But there will come a time when the ordinary star 
circling the black hole will begin to turn into a red giant or supergiant. 
Its atmosphere will leak onto the black hole. Just as it does when a star's 
atmosphere falls on to a neutron star, the compressed gas will be heated 
to temperatures of millions of degrees centigrade and will radiate 
X-rays. To find a black hole, said Shklovsky, look among the X-ray 
stars. But how can we distinguish X-ray emissions from a black hole 
from those from a neutron star ? 

The major fact which would distinguish the black hole would be its 
mass, perturbing the companion star by the force of gravity. If the mass 
of such an X-ray source, measured by the size of the perturbation of the 
companion star, was larger than the extreme upper limit to the allowable 
masses of a neutron star, then the X-ray source might be a black hole. 

The first such X-ray source known, the invisible companion to a large 
but ordinary star, is Cygnus x-i. First observed by rocket- and balloon- 





ttD 226868 

In Cygnus X-i a large blue supergiant star, HD 226868, may be overflowing onto 
a companion black hole. Gas from its atmosphere leaks towards the black hole, 
encircling it before falling in and being heated so as to emit X-rays. 

borne X-ray telescopes in the mid 1960s, Cygnus x-i was one of the 
first X-ray stars studied by the Uhuru satellite in 1971. With this 
satellite a group of X-ray astronomers, led by R. Giacconi and H. 
Gursky, located Cygnus x-i in a small area of the sky in which radio 
astronomers (L. Braes and G. Wiley in Holland and R. M. Hjellming 
and C. Wade in the U.S.) had spotted a radio star which had not been 
visible before. At the same time that the radio star turned on, Cygnus x-i 
changed its X-ray character, proving that the X-ray star and radio star 
were one and the same object. 

At the same position of the radio star was a visible star, numbered 
226868 in the Henry Draper catalog (hd). hd 226868 was immediately 
found to be a hot supergiant star, about 30 solar masses, but not peculiar 
in any way. The blue supergiant was so normal that two groups of 
astronomers who studied hd 226868 concluded that it was also a red 
herring, and nothing to do with Cygnus x-i. But two other groups of 



astronomers in England and Canada kept observing the star to see 
whether it changed at all. Tom Bolton at the University of Toronto and 
Louise Webster and Paul Murdin at the Royal Greenwich Observatory 
simultaneously published their findings that it did. The star had a cycle 
of radial velocity change which lasted 5.6 days as it orbited an invisible 
star, alternately approaching and receding from Earth so that the lines 
in its spectrum were blue- and red-shifted by the Doppler effect. 

It was not possible immediately to say precisely what the mass of the 
invisible companion was, since the inclination of the orbit of the binary 
star was unknown and the astronomers could not tell whether they saw 
the full motion (orbit seen edge-on) or a small part (orbit seen nearly 
face-on), but clearly the mass of the invisible companion had to be at 
least six times that of the Sun to swing hd 226868 as it did. This mini- 
mum is more than is possible if the invisible companion were a neutron 
star, being more than the critical mass of neutron stars, namely 3.2 solar 

Thus Cygnus x-i fits in detail the scenario for the discovery of a 
black hole in a binary star system, outlined by I. Shklovsky in 1967 as a 
solution before the existence of the problem. 
table vii The velocity of escape 


1 1 km/s 









Typical stars : 

blue dwarf 


red dwarf 


blue supergiant 


red supergiant 


Collapsed stars: 

white dwarf 


neutron star 


(1 km/s = 224omiles/hr; speed of light = 300,000 km/s.) 


This is not to say that Cygnus x-i must be a black hole. Perhaps the 
X-ray emitter is a neutron star (of mass i solar mass) orbiting a dim 
normal star (of, say, five solar masses) which itself orbits the bright and 
visible supergiant. But the black hole explanation accounts for the 
observed facts and has the attraction that it was proposed before the 
facts were known; it does not suffer from the suspicion that it has been 
patched to fit in. Perhaps a black hole, exotic though it may be, has 
indeed been discovered lurking in Cygnus x-i, formed by a long-past 
supernova and enabled to shine by devouring its companion. 


XV Final chapter 

Some say the world will end in fire, 
Some say in ice, 
From what I've tasted of desire 
I hold with those who favor fire, 
But if it had to perish twice, 
I think I know enough of hate 
To say that for destruction ice 
Is also great 
And would suffice. 

Robert Frost 


.he escape velocity is a concept which can be applied to 
the constituents of the whole Universe as well as to a star or planet 
within it. All matter in the Universe was subject to the explosion of the 
Big Bang and may have been given the velocity of escape from the 
gravitational pull of the rest of the Universe. If this is so, then the energy 
of motion of the Universe, its kinetic energy, is larger than its gravita- 
tional energy and the explosion which began the Universe will never 
end: the Universe will continue to disperse for ever. If, on the other 
hand, the Big Bang was not powerful enough to overcome the mutual 
gravitational attraction of all parts of the Universe, the explosion will 
eventually coast to a halt and the Universe will collapse; when it gets 
small enough, it may re-explode and bounce, oscillating indefinitely. 

In the technical jargon on the subject, if the gravitational energy of 
the Universe exceeds the kinetic energy, the Universe is closed and will 
collapse; if its gravitational energy is less than its kinetic energy, the 
Universe is open and will expand forever. 



There are two direct lines of attack on the problem of deciding 
between these possibilities. The first consists of looking back in time at 
distant galaxies so far away that they represent the Universe as it was a 
significantly long time ago, and trying to see what the expansion rate of 
the Universe was then. The expansion rate may be slowing down so 
quickly that we can tell whether the Universe will decelerate to a stop 
and collapse. 

To make this method work, astronomers must first measure the cosmic 
expansion rate of a distant group of galaxies and then determine their 
distance. It is here that the difficulties arise. If galaxies were all of 
known intrinsic brightness, like cepheid variable stars, cosmologists 
could use their apparent brightness as a distance measurement. In the 
past 20 years, several attempts using this method to determine whether 
the Universe will expand forever have marginally favored the result that 
it is closed and will ultimately collapse. But soon after the Big Bang, the 
galaxies all formed at about the same time, so that distant galaxies are 
younger than nearby ones (because as we look farther away we look back 
in time). If younger galaxies are brighter than older ones (because they 
contain larger numbers of bright stars), they will seem nearer than they 
really are. Hence we will be measuring the expansion rate for distant 
galaxies as though they were nearby, and overestimating the amount of 
deceleration of the Universe. 

These results thus seem to be overestimates of the deceleration and 
may be biased in saying that the Universe is closed. 

Another way in which the distance of far galaxies can be obtained is 
to look at their angular size. This method, applied to clusters of galaxies 
and to radio galaxies, has given values of the deceleration which are on 
the borderline between closed and open Universes and are tantalizingly 
equivocal. They marginally favor the open Universe. 

Perhaps a better method of determining whether the Universe is open 
or closed is to attempt directly to estimate the kinetic and gravitational 
energy of the Universe to see which is bigger. 

In attempting to add up all the mass in the Universe item by item to 



calculate its gravitational energy, astronomers have come up against the 
problem that they simply do not know enough about what kind of 
material predominates in the Universe. Most of the mass of which they 
are cognizant is in the form of galaxies, and the mass of an average 
galaxy can be measured in two ways. Astronomers can look at the speed 
with which stars in the outer parts of a galaxy orbit its center, and esti- 
mate the mass required to deflect stars by the amount observed (just as 
the mass of the companion star to Cygnus x-i has been estimated by 
looking at its effect on the visible star hd 226868). Alternatively, they 
can look at the speeds with which individual galaxies deflect each other 
when they are situated in a cluster of galaxies. 

The former method suffers from the disadvantage that if there is a 
halo of very faint, undetectable stars surrounding the galaxy, these have 
no effect on the motion of the visible stars nearer the galaxy's center and 
they remain undetected. Possibly this is why the two methods for esti- 
mating the mass of an average galaxy give answers differing by a factor 
of 100 ! Somehow, astronomers may be missing 99 per cent of the matter 
in an average galaxy! 

What form this missing matter takes has been the subject of much 
speculation. None of the observations exclude the possibility that 
between or around galaxies lie enormous masses of black holes, faint red 
dwarf stars, rocks or hot gas (at temperatures of around a million 
degrees). There is not much cold hydrogen, for this would be seen to 
play a larger part in absorbing the light from distant quasars than it in 
fact does. Recent discoveries by the Uhuru X-ray satellite of X-rays from 
the vicinity of clusters of galaxies has shown that between galaxies in the 
clusters there does exist very hot gas (at 100 million degrees), but 
probably not in such abundance that it can close the Universe. 

Instead of looking at the components of the Universe item by item 
and adding them all up to determine the gravitational energy of the 
Universe, Allan Sandage has attacked the problem by looking at how 
faithfully the nearby galaxies follow the Hubble law, that their redshifts 
are proportional to their distances from us. He argues that where galaxies 



do not follow this law closely, the departures from the law are caused by 
local clumps of matter (other galaxies, other clusters of galaxies or 
whatever) and the amount by which they are deflected from the Hubble 
law tells how much matter is deflecting them. Sandage obtains a result 
which he tersely summarizes: "Taken at face value these values suggest 

that (i) the deceleration is almost negligible (2) the Universe is open, 

and (3) the expansion will not reverse." 

If Sandage is right, the Galaxy will become increasingly isolated from 
its neighbors as they recede from it. The Galaxy itself will ultimately 
cease to shine. Already a significant fraction of its mass is locked up in 
dark stars — white dwarfs and the end products of supernovae : black 
holes and neutron stars. An increasing proportion of its gas will have 
processed through stars and increasing amounts of metals will be thrown 
back into the interstellar medium by supernovae. When the gas gets too 
metal-polluted, stars which form from it will not be able to shine. 

Supernovae not only mark the death of individual stars, they hasten 
the aging of our Galaxy, possibly towards a dark, cold and lonely death 
as it finds itself alone in the Universe. 

At a conference in Cracow in 1973, John Wheeler conducted an opinion 
poll of the assembled cosmologists on the question of whether they 
thought the Universe was closed or open. Of course, truth is not decided 
by democratic vote — the result of the poll only gives an indication of 
what most informed people think. Most cosmologists put themselves into 
the "don't know" category, and were prepared to wait for more solid 
evidence before pronouncing on the subject. We apparently shall not for 
a while be able to read the final chapter in the life of our Galaxy, 
although it is at this moment being written sentence by sentence among 
the stars, and punctuated by supernovae. 



Books at a similar level to this one, containing relevant material 

N. Calder, The Violent Universe. BBC publications. London, 1968. 
D. Bergamini, The Universe. Time-Life Books. 1964. 
F. Hoyle, Astronomy. Macdonald. London, 1962. 
J. S. Glasby, Variable Stars. Constable. London, 1968. 
T. Weekes, High Energy Astrophysics. Chapman and Hall. London, 1968. 
I. Ridpath (ed.), Illustrated Encyclopedia of Astronomy and Space. T. Y. 
Crowell. N.Y., 1976. 

Books about supernovae and their remnants available only in specialist 
libraries, or through the interlibrary loan system 

Krishna, M. V. Apparao, The Crab Nebula. Astrophysics & Space Science, 

vol. 25, p. 3, 1973. 
P. J. Bancazio and A. G. W. Cameron (ed.), Supernovae and their remnants. 

Gordon and Breach. N.Y., 1969. 
D. H. Clark and F. Stephenson, The Historical Supernovae. Pergamon. London 

and N.Y., 1977. 
C. B. Cosmovici (ed.), Supernovae and Supernovae Remnants. Reidel. Dordrecht, 

R. D. Davies and F. G. Smith (ed.), The Crab Nebula. Reidel. Dordrecht, 1971. 
Flagstaff Symposium on the Crab Nebula, Publications of the Astronomical 

Society of the Pacific, vol. 82, p. 375, 1970. 
R. N. Manchester and J. H. Taylor, Pulsars. W. H. Freeman. San Francisco, 

I. S. Shklovsky, Supernovae. Wiley. London, 1968. 
I. S. Shklovsky, Cosmic Radio Waves. Harvard. Cambridge, i960. 



Abt, Helmut, 10 

Adams, WS., 198 

Alib.Ridwan, 15 

Alpha- beta-gamma theory; 164 



Antimatter, 159 

Aquinas, Thomas, 25 

Archaeological dating, 182 

Aristotelian universe, see Ptolemaic cosmology 

Aristotle, 25 

Arizona cave drawings, 9-11 

Astrology, 44 

B 2 FH paper, 168 

Baade, Walter, 18, 56, 86, 122, 131 

Balloons, use in astronomy, 93 


Beta-decay, 155 

Bethe, Hans, 165, 167 

Bethlehem, Star of, 21 

Bevisjohn, 67 

Bevis Nebula, wCrab Nebula 

Big Bang theory, 164 

Binary stars, 186; eclipsing, 191; single-line 

spectroscopic, 202 
Blaauw, Adriaan, 188 
Black dwarf, 147 
Black hole, 199 ff 
Bolton, John, 90 
Bolton, Tom, 204 

Brahe, Tycho, 23, 33 ff; mentioned by Dryden, 46 
Brahe's supernova, 134; magnitudes, 136 
Brightness, measurement of, 134 
Brown, Robert Hanbury, 92 
Browne, Sir Thomas, 44 
Bucha, V, 183 

Burbidge, Geoffrey, 5 ; and Margaret, 168 
Bursters, 194 

Calendar, problems of, 22 

Cassiopeia A, 122 

Chandrasekhar, S., 147 

Chandrasekhar Limit, 147, 197 

Chang Te-hsiang, 7 

Chinese astronomy, 6-7 

Christ, birth of, 21 ff 

Clark, David H., 13 

Clarke, Arthur C, 23 

Clerke, Agnes, 51 

Conservation of angular momentum, 151 

Copernicus, 31 

Corbet, Richard, Bishop of Oxford, 44 

Cosmic radiation, 4, 179; and evolution, 184 

Crab Nebula, 3, 5, 66 ff, 91; Doppler effect in, 80; 

expansion rate of, 87; first photograph of, 71; 

pulsar, 110 ff; radio map of, 92 
Curtis, H.D., 54 
CygnusX- 1,203-5 

Dating, problems of, 22 
Degeneracy, 104 
Digges, Thomas, 27, 32 
Dinosaurs, 184 
Dispersion, 103 
Distance, astronomical, 85 
Dodd, William W, 122 
Donnejohn, 28, 45 
Doppler effect/shift, 77, 101, 192 
Dover Heights Telescope, 90 
Dreyer, 69 
Dryden, John, 46 
Duncanjohn, 82 
Dust correction, 140 
DuyvendakJJ.L, 83 

E-process, 168 

Eclipsing binary stars, 191 

Eddington, Arthur Stanley, 198 

Electron, 177 ff 

Elements, 163; creation of (table), 176 

Energy of stars, 145 ff 

Epicyclic movement of planets, 31 

Escape velocity, 196 ff; (table), 204 

Evolution, effect of cosmic radiation on, 184 

Extraterrestrial life, 101 

Fabricius, David, 41 
Fermi, Enrico, 156 
"Fossil" nebulae, 130 
Fowler, Willy, 168 

Galaxies, shape of, 142; death of, 209 

Galileo Galilei, 42 ff; and the telescope, 47 

Gamow, George, 165 

Gardner, Frank, 18 

Glitch, 153 

Globular clusters, 54 

Gold, in the Sun, 163 

Gold, Thomas, 114, 151 

Gravitational redshift, 198 

Gravity, 195 ff 

Guest stars, 6 ff 

Gum, Colin, 125 

Gum Nebula, 125 ff 

Hartwig, E., 50 



Hcpidannus, 16 

Herschel, William, 69 

Hewish, Antonv, 98 


Hipparchus, 29 

Hovle, Fred, 11, 107, 168; as novelist, 106 

Hubble, Edwin, 55, 82, 139, 142 

Hubble redshift relation, 139 

"Hubble time," 82 

Huggins, William, 72 

Humason, Milton, 131 

Interferometer, 90 
Ionization, 178 

Jansky, Karl, 88 

Kepler, Johannes, 23, 38 fT; on Brahe, 34 
Kepler's Supernova, magnitudes, 136 

Lampland, CO., 81, 86 
I-aplace, Pierre-Simon, 50 
Lassell, William, 70 
l.clande, 67 
LGM- 1,101 
Lick Observatory, 112 
Lipperhey, Hans, 47 
Lone star system, 186 
Lovell, Bernard, 107 
Lundmark, K., 82 
Lyne, Andrew, 60 

Ml, see Crab Nebula 

Magnitude scale, 134 

Mass, 143 

Mastlin, Michael, 32 

Maurolvco, Francesco, 32 


Measurement, Brahe's method, 36 

Measurement of distance in space, 103 

Merrill, Paul, 167 

Messier, Charles, 67 

Metal rich/poor stars, 173 

Meteor, 28 

Michanowsky, George, 189 

Middleditchjohn, 192 

Miller, Bill, 10 

Miller, Joe, 112 


Mills, Bernard, 92 

Milne, Douglas, 18, 130 

Mnkowski, Rudolph, 86, 122, 131, 141 

More, Henry, 44 

Mugnoz, Hieronymus, 32 

Multiple star system, 186 

Murdin, Paul, 204 

Nebula, 3; catalogue of, 68; "fossil," 130; 
planetary, 148; see also Crab Nebula 

Nebulium, 75 

Nelsonjerrv, 192 

Neutrino, 155 ff; detector, 157 

Neutron, 156 

Neutron star, 3, 105, 149-150; X-radiation from, 191 

Newton, Isaac, 72 

Nobel Prize, 107 

Nova Persei 1901, 53 

Novae, 3; creation of, 193; recurrent, 193 

Nucleocosmochronology, 170 

Observation, astronomical, 66-67 
Oort, Jan, 83 
Origin of planets, 50 
Osiander, Andrew, 31 

P-process, 169 

Pacini, F, 114 

Palomar Supernova Search, 131 

Parallax, 36 


Photometer, 134 

Pickering, EC, 72, 75 

Planetary nebula, 148 

Plasma, 103 

Pontecorvo, B., 157 

Proper motion, 86 

Protostars, 144 

Ptolemaic cosmology, 27; rejection of, 41-2; 

compared to Copernican, 48 
Ptolemv of Alexandria, 25 
Pulsars, 97 ff; as lone stars, 188; birth rate, 60; 

definition of, 4; magnetic field of, 152; 

optical observation of, 112; shortest 

period, 109 - 110; size, 104; slowest pulse 
■ rate, 188 

Quasars, 97; redshift in, 199 

R-process, 169 

Radiation, creation of, 121; synchrotron radiation, 95; 

see also X-radiation 
Radio astronomy, early, 88 
Radio star, 91 

Radio telescope, see Telescope 
Radiocarbon dating, 182 
Reber, Grote, 88 
Red giant, 146; in binary systems, 193; presence of 

techtenium, 167; gravitational/Doppler confusion, 

Reinfenstein, EC, 109 
Roberts, Isaac, 71 
Roche lobes, 191 
Rockets, use in astronomy, 93 
Rosenberg, Ivan, 123 
Rosse, Earl of, 69 
Ruflin, R, 201 
"Runaway" stars, 188 



Ryle, Martin, 107,122 

S Andromedae, 52 ff 

S-process, 169 

Salpeter, E.E., 168 

Sandage, Allan, 208 

Sargent, Wallace, 58 

Satellites in astronomy, 93 

Schmidt, Bernhard, 57 

Scintillation, 98 

Searle, Leonard, 58 

Shapley, Harlow, 54 



Smith, F. Graham, 122 

Solar system, creation of, 170 

Solar wind, 190 

Spallation, 181 

Spectroscope, 72 

Spenser, Edmund, 46 


Stanley, G.J., 90 

Stars, colour of, 73, 75; life-cycle of, 1 - 2, 144; 

lone/multiple states, 186; rotation speed 

of, 150 ff; speed of, 192; system of naming, 12 
Starquake, 153 
Static electricity, 177 
Steady State theory, 168 
Stephenson, E Richard, 13 
Steward Observatory, Arizona, 112 
Sun, 145 

Supernova of 1054, 8 ff 
Supernova remnants, 3, 118 ff; list of, 123; 

shape of, 121 
Supernovae, 1 ff; appearance of, 64; astrological 

view of, 44; first statement of theory, 55; 

function in scattering elements, 169; list of 

historical, 14; neutrinos from, 158 ff; 

radiation from, 64; Renaissance, 25 ff; 

searches for, 57-9; types of, 132 ff 
Synchrotron radiation, 95 
Synchrotrons, 75 

Taurus A, see Crab Nebula 

Telescope, 68; Anglo- Australian 154" 116; 4'/2acre 

Cambridge, 97; Hey's, 90; invention and 

development, 3, 47; Mount Wilson 100", 118; 

radio, 5 ; radio - interferometer type, 90; 

Rosse's, 70; rotating shutter, 112; Schmidt 

camera telescope, 57 
Theory of relativity, 197 
Thomson, J.J., 178 
Tinsley, Beatrice, 175 
Triple-alpha process, 146 
2001 (Clarke), 185 
Tycho, see Brahe 

Universe, age.of, 82; future of, 206 ff; material 
nature of, 208 

Urea process, 159 

Van den Bergh, Sidney, 18, 65, 122 

Van Weizsacker, 167 

Vela pulsar, optical observation of, 116 

Vela X, 109, 130 

Velocity of escape, 196 ff; (table), 204 

Verne, Jules, 191 

Very, FW, 53 

Wampler, Joe, 112 

Webster, Louise, 204 

Wheeler, J.A, 201 

White dwarf, 104, 147; in binary system, 193 

Wilson, AS, 92 

Winlock, Joseph, 72, 75 

Wulf, 178 

X-process, 180 

X- radiation, 92; from Crab pulsar, 115 

X-ray pulsar, 190 

X-ray star, 5; binary (table), 194 

Yang Wei-te, 7 
Ylem, 165 

Zeta Tauri, 6